Variability of aerosol and cloud optical properties and their effect on the transfer of solar irradiance in the atmosphere

Transcription

1 UNIVERSITY OF PATRAS PHYSICS DEPARTMENT LABORATORY OF ATMOSPHERIC PHYSICS Doctoral Thesis Variability of aerosol and cloud optical properties and their effect on the transfer of solar irradiance in the atmosphere (Διακυμάνσεις των οπτικών ιδιοτήτων των αιωρούμενων σωματιδίων και των νεφών και η επίδραση τους στο ισοζύγιο της ηλιακής ακτινοβολίας στην ατμόσφαιρα) Nikitidou Efterpi MSc, Physicist Patras, Greece, 2013

2 2

3 3 Στην οικογένεια μου

4 4

5 5 Πρόλογος Η παρούσα διδακτορική διατριβή εκπονήθηκε στο Εργαστήριο Φυσικής της Ατμόσφαιρας, του Τμήματος Φυσικής του Πανεπιστημίου Πατρών, υπό την επίβλεψη του Επίκουρου Καθηγητή Ανδρέα Καζαντζίδη. Θα ήθελα να ευχαριστήσω, από τα βάθη της καρδιάς μου, τον επιβλέποντα μου, για όλα όσα μου προσέφερε αυτά τα χρόνια, που έχω τη χαρά να δουλεύω μαζί του. Αποτέλεσε για μένα ένα σταθερό σημείο έμπνευσης και εμπιστοσύνης, τόσο σε ερευνητικό επίπεδο, όσο και σε ανθρώπινο, κάτι που, ως ιδιαίτερα σπάνιο και ιδιαίτερο, εκτιμώ περισσότερο από όλα. Η διαρκής συμμετοχή και συμπαράσταση του, καθώς και το απίστευτο χιούμορ, που τον διακατέχει, ήταν καταλυτικά στοιχεία, χωρίς τα οποία δεν θα έβρισκα τη δύναμη να ολοκληρώσω αυτό το εκπόνημα. Με τη βοήθεια του, πραγματοποίησα βήματα που μου έδωσαν πολύτιμες εμπειρίες για τη συνέχεια και εύχομαι η συνεργασία μας να συνεχιστεί και στο μέλλον. Θα ήθελα να ευχαριστήσω και τα άλλα δύο μέλη της τριμελούς επιτροπής μου, τον Αναπληρωτή Καθηγητή Αθανάσιο Αργυρίου και τον Καθηγητή Αλκιβιάδη Μπάη, για τη βοήθεια που μου προσέφεραν, καθώς και τα υπόλοιπα μέλη της επταμελούς επιτροπής. Δεν θα μπορούσα να μην αναφερθώ σε όλους τους καθηγητές και καθηγήτριες μου, του Μεταπτυχιακού προγράμματος Φυσικής Περιβάλλοντος, του Αριστοτελείου Πανεπιστημίου Θεσσαλονίκης, που μου έδωσαν τις βάσεις για να προχωρήσω στο επόμενο βήμα. Σημαντική ήταν και η συνεργασία με τον Dr Hugo De Backer και την ερευνήτρια Veerle De Bock, του Βασιλικού Μετεωρολογικού Ινστιτούτο του Βελγίου, που με δέχτηκαν και με βοήθησαν, το διάστημα που βρέθηκα κοντά τους και συντέλεσαν στην εκπόνηση ενός κομματιού αυτής της διατριβής. Ευχαριστώ πολύ τον υποψήφιο διδάκτορα, Παναγιώτη Τζουμανίκα, για την άψογη συνεργασία μας και τον υποψήφιο διδάκτορα, Βασίλη Σαλαμαλίκη, για όλα όσα περάσαμε στο εργαστήριο, τη δουλειά και το γέλιο που κάναμε, που έκαναν τον καιρό που πέρασα στην Πάτρα, απολαυστικό! Η παρούσα διατριβή δεν θα μπορούσε να ολοκληρωθεί χωρίς τη συμπαράσταση των στενών μου φίλων, που με στήριξαν και με βοήθησαν, μένοντας δίπλα μου τόσο στις χαρές όσο και τις στιγμές που οι

6 καταστάσεις δυσκόλευαν. Η παρουσία και η στήριξη τους με βοήθησαν κάθε στιγμή και τους χρωστάω ένα μεγάλο ευχαριστώ. 6 Το μεγαλύτερο ευχαριστώ το κράτησα στο τέλος, για την οικογένεια μου, που βρίσκεται πάντα δίπλα μου και με υποστηρίζει όλα αυτά τα χρόνια, σε ό,τι αποφάσεις παίρνω και ό,τι δρόμους επιλέγω να διανύσω. Οι θυσίες τους, η αγάπη και η παντοτινή στήριξη τους είναι κάτι, χωρίς το οποίο δεν θα είχα καταφέρει τίποτα. Στους γονείς μου, Γιούλη και Μάκη, και την αδελφή μου, Νικολέτα, αφιερώνω αυτή τη διατριβή.

7 7 Περίληψη Η παρούσα διατριβή ασχολείται με τις οπτικές ιδιότητες των αιωρούμενων σωματιδίων και των νεφών και τις επιδράσεις που αυτές έχουν στη διάδοση της ηλιακής ακτινοβολίας στην ατμόσφαιρα. Το πρώτο κεφάλαιο παρέχει μια σύντομη περιγραφή των βασικών αρχών που διέπουν τη διάδοση της ηλιακής ακτινοβολίας. Η θεωρία της διάδοσης της ακτινοβολίας περιγράφεται, μαζί με διάφορες προσεγγίσεις, που χρησιμοποιούνται για τη λύση συγκεκριμένων προβλημάτων στις ατμοσφαιρικές επιστήμες. Τα συστατικά της ατμόσφαιρας, που είναι άμεσου ενδιαφέροντος σε αυτήν τη διατριβή, τα αιωρούμενα σωματίδια και τα νέφη, περιγράφονται, με βάση τους τύπους τους και τις οπτικές τους ιδιότητες, ενώ περιγράφονται ακόμα οι βασικές αρχές της σκέδασης και της απορρόφησης, μέσω των οποίων επηρρεάζουν τη διάδοση της ηλιακής ακτινοβολίας. Το δεύτερο κεφάλαιο παρέχει μια περιγραφή των επίγειων δικτύων, μοντέλων και δορυφορικών οργάνων, των οποίων τα δεδομένα χρησιμοποιήθηκαν, για τη διεκπαιρέωση αυτής της διατριβής, μαζί με την περιγραφή του μοντέλου διάδοσης της ακτινοβολίας, που χρησιμοποιήθηκε για τους θεωρητικούς υπολογισμούς. Το τρίτο κεφάλαιο επικεντρώνεται στις οπτικές ιδιότητες των αιωρούμενων σωματιδίων, στο υπεριώδες και ορατό κομμάτι του ηλεκτρομαγνητικού φάσματος, στην Μεσόγειο. Τρεις ξεχωριστές βάσεις δεδομένων, από επίγειους σταθμούς, μοντέλα και δορυφορικά όργανα, χρησιμοποιούνται για τον υπολογισμό της ακτινοβολίας στο υπεριώδες και ορατό, σε οχτώ σταθμούς στην περιοχή της Μεσογείου. Χρησιμοποιούνται δεδομένα από το AERONET, το AeroCom και το MODIS και μελετούνται οι διαφορές στις υπολογιζόμενες, από το μοντέλο, ακτινοβολίες, οι οποίες προκύπτουν από τις διαφορές στις οπτικές ιδιότητες των αιωρούμενων σωματιδίων, που παρέχονται από κάθε βάση δεδομένων. Οι ακτινοβολίες υπολογίζονται με το μοντέλο διάδοσης ακτινοβολίας libradtran. Τα δεδομένα του MODIS βρίσκονται σε καλύτερη συμφωνία με αυτά του AERONET, με τη μέγιστη διαφορά στο οπτικό βάθος, στα 550 nm, να είναι ίση με 0.09, ενώ οι αντίστοιχες διαφορές με το AeroCom υπολογίζονται στα 0.25 και 0.15, για το υπεριώδες και ορατό αντίστοιχα. Ως αποτέλεσμα, οι απόλυτες διαφορές στις υπολογιζόμενες ακτινοβολίες, μεταξύ AERONET και MODIS υπολογίζονται γύρω στο 6%, ενώ αυτές που αφορούν την κλιματολογία

8 AeroCom φτάνουν το 12%. Οι μεγαλύτερες διαφορές αφορούν περιοχές που επηρεάζονται από σωματίδια ερημικής σκόνης. 8 Στο τέταρτο κεφάλαιο, η άμεση επίδραση των αιωρούμενων σωματιδίων, στην υπεριώδη ακτινοβολία, μελετάται, για μια τυπική περιοχή της Δυτικής Ευρώπης. Μετρήσεις από ένα όργανο Brewer, που λειτουργεί στην περιοχή και θεωρητικοί υπολογισμοί με το μοντέλο libradtran, χρησιμοποιούνται για τον υπολογισμό της ικανότητας κλιματικού εξαναγκασμού των αιωρούμενων σωματιδίων, στο φάσμα nm και στο UV-B φάσμα των nm. Μετρήσεις από το όργανο και θεωρητικοί υπολογισμοί, χρησιμοποιούνται στη συνέχεια για τον υπολογισμό της ανακλαστικότητας μεμονωμένης σκέδασης των αιωρούμενων σωματιδίων, σε χαμηλά UV-A και σε UV-B μήκη κύματος. Στο φάσμα nm, οι μεγαλύτερες τιμές της ικανότητας κλιματικού εξαναγκασμού, παρατηρούνται στις 30 o (-6.9 ± 0.9 W/m 2 ), ενώ στις 60 o οι τιμές είναι σχεδόν 2.5 φορές χαμηλότερες (-2.7 ±0.1 W/m 2 ). Στο UV-B κομμάτι του φάσματος ( nm), οι αντίστοιχες τιμές στις 60 o και 30 o υπολογίζονται ίσες με ±0.005 W/m 2 και ±0.04 W/m 2. Συγκρίνοντας τις τιμές που προκύπτουν για την ανακλαστικότητα μεμονωμένης σκέδασης στα 320 nm, με αυτές από το γειτονικό CIMEL στα 440 nm, προκύπτει πολύ καλή συμφωνία (±0.01). Το πέμπτο κεφάλαιο, επικεντρώνεται στην επίδραση των αιωρούμενων σωματιδίων στην άμεση ηλιακή ακτινοβολία, σε επίπεδο κάθετο στην κατεύθυνση της ακτινοβολίας, στην περιοχή της Ευρώπης. Δεδομένα από το MODIS, το AERONET και θεωρητικοί υπολογισμοί με το μοντέλο SBDART, χρησιμοποιούνται για τον υπολογισμό της ημερήσιας ποσότητας άμεσης ηλιακής ακτινοβολίας, στην Ευρώπη, με χωρική ανάλυση 1 x1 για μια χρονική περίοδο 13 ετών. Ο κλιματικός εξαναγκασμός, υπό ανέφελο ουρανό, των αιωρούμενων σωματιδίων και πιθανές μεταβολές στην ληφθείσα άμεση ακτινοβολία κατά τη διάρκεια αυτής της περιόδου, μελετούνται. Οι επιδράσεις των αιωρούμενων σωματιδίων είναι σημαντικές σε περιοχές που επηρεάζονται από σωματίδια ερημικής σκόνης και περιοχές με έντονη ανθρωπογενή δραστηριότητα, όπως η Μεσόγειος και η κοιλάδα του Πάδου στην Ιταλία. Σε αυτές τις περιοχές η μείωση της ακτινοβολίας, λόγω αιωρούμενων σωματιδίων, φτάνει, το Μάιο, το 35% και 35-45%, που αντιστοιχεί σε 4 και kwh/m 2 την ημέρα. Η ληφθείσα άμεση ακτινοβολία έχει αυξηθεί κατά τα

9 9 τελευταία χρόνια, λόγω ελάττωσης της συγκέντρωσης των αιωρούμενων σωματιδίων σε πολλά μέρη της Ευρώπης. Οι μεγαλύτερες αυξήσεις κυμαίνονται μεταξύ 6 και 12%, ποσοστό που αντιστοιχεί σε 0.5 με 1.25 kwh/m 2 την ημέρα. Στο έκτο κεφάλαιο, αυτής της διατριβής, μελετάται ο υπολογισμός της ηλιακής ακτινοβολίας στο έδαφος, χρησιμοποιώντας δορυφορικά δεδομένα για την επίδραση των νεφών. Το όργανο SEVIRI, στους δορυφόρους MSG, χρησιμοποιείται για την παροχή δεδομένων σχετικά με το συντελεστή επίδρασης των νεφών. Τα δεδομένα αυτά, μαζί με θεωρητικούς υπολογισμούς με το μοντέλο libradtran, χρησιμοποιούνται για τον υπολογισμό της ολικής ηλιακής ακτινοβολίας, σε οριζόντια επιφάνεια και σε επιφάνεια υπό κλίση, καθώς και τον υπολογισμό της άμεσης συνιστώσας σε επιφάνεια κάθετη στη διεύθυνση της ακτινοβολίας. Η μελέτη πραγματοποιείται για της περιοχή της Ελλάδας και αποτελεί κομμάτι του Ελληνικού Δικτύου Ηλιακής Ενέργειας, που έχει αναπτυχθεί για την υποστήριξη εφαρμογών και συστημάτων ηλιακής ενέργειας. Υπολογίζονται οι ημερήσιες ποσότητες ακτινοβολίας, οι μηνιαίες και οι ετήσιες τιμές, για μια περίοδο 11 ετών, καθώς και η μηνιαία κλιματολογία που προκύπτει για αυτήν την περίοδο. Η σύγκριση των αποτελεσμάτων με μετρήσεις από επίγειους σταθμούς, δίνει πολύ καλή συμφωνία, ενώ οι μεγαλύτερες διαφορές παρατηρούνται σε περιπτώσεις πολύ πυκνών νεφών. Η ηλιακή ακτινοβολία που συλλέγεται σε κεκλιμένη επιφάνεια, παρέχει % μεγαλύτερα ποσά, σε σχέση με αυτήν που παρέχουν οριζόντιες επιφάνειες συλλογής. Στην Ελλάδα, τα μεγαλύτερα μηνιαία ποσά ηλιακής ενέργειας, παρατηρούνται κατά τους θερινούς μήνες, στη Νότια Πελοπόννησο, την Κρήτη και τις Κυκλάδες και ξεπερνούν τις 250 kwh/m 2.

10 10 Abstract This thesis is focused on the aerosols and clouds optical properties and the effects that these parameters have on the solar radiation transfer in the atmosphere. The first chapter provides a brief description of the basic concepts of radiative transfer. The radiative transfer theory is described, along with various approximations, used to address specific atmospheric transfer problems. The atmospheric constituents, which are of interest of this thesis, aerosols and clouds, are described, in terms of their types and radiative properties and the main aspects of the scattering and absorption that they induce on the solar radiation, are provided. The second chapter provides a description of the networks, models and satellite instruments, whose data were used in this thesis, along with a description of the radiative transfer model, used for the simulations. Chapter three focuses on the aerosol optical properties in the ultraviolet and visible wavelength ranges, in the Mediterranean. Three datasets, from ground-based stations, global aerosol models and satellite instruments, are used to simulate the corresponding irradiances in the UV and VIS, in eight stations in the Mediterranean basin. Data from AERONET, AeroCom and MODIS are used and the differences on the modeled irradiances, which arise from the different aerosol optical properties provided by each dataset, are examined. The irradiance simulations are performed with the libradtran radiative transfer model. The MODIS aerosol optical depth climatology shows better agreement with AERONET data. The highest difference in the monthly average values is equal to 0.09 at 550nm, while the differences between the AERONET and the AeroCom climatologies reach 0.25 and 0.15 in the UV and VIS wavelengths respectively. As a result, the AERONET modeled VIS and UV irradiances are closer to MODIS, with the absolute differences in average values reaching 6%, while absolute differences with AeroCom irradiances can reach up to 12%. The differences are higher in areas affected by desert dust aerosols. In chapter four, the aerosol direct effect on the UV solar irradiance, is examined, at a typical West European site. Measurements from a Brewer instrument, operating at the site, are used, along with model simulations, provided from libradtran, to estimate the aerosol forcing efficiency in the nm spectral region and in the UV-B region of nm. Instrument measurements and model calculations are

11 11 subsequently used to derive the aerosol single scattering albedo at low UV-A and at UV-B wavelengths. In the nm spectral region, the highest values were revealed at 30 o (-6.9 ± 0.9 W/m 2 ), while at 60 o the RFE was almost 2.5 times lower (-2.7 ±0.1 W/m 2 ). In the UV-B region ( nm), the RFE value at 60 o and 30 o was estimated to be equal to ±0.005 W/m 2 and ±0.04 W/m 2, respectively. The estimated monthly averages of the Brewer single scattering albedo at 320 nm are in very close agreement (within ±0.01) with measurements at 440nm from a collocated CIMEL sunphotometer. Chapter five focuses on the aerosol effect on the Direct Normal Irradiance, in the area of Europe. Data from the MODIS satellite instrument, AERONET network and model simulations with SBDART, are used to calculate the daily amount of Direct Normal Irradiance received in the European continent, with a spatial resolution of 1 x1, for a 13-year period. The clear-sky aerosol radiative forcing is calculated and possible variations in the received Direct Normal Irradiance, during the 13-year studied period, are examined. The clear-sky aerosol radiative forcing on Direct Normal Irradiance is high in areas influenced by desert dust and intense anthropogenic activities, such as the Mediterranean basin and the Po Valley in Italy. In May, the attenuation from aerosols, over these areas, can reach values up to 35% and 35-45%, which corresponds to 4 and kwh/m 2 per day, respectively. The Direct Normal Irradiance received, seems to have increased during the recent period, due to the decreasing trend of aerosol load, over many parts of Europe. The largest increases are around 6 to 12%, which correspond to an amount of 0.5 to 1.25 more kwh/m 2 received per day. Finally, chapter six focuses on the retrieval of solar irradiance on the ground, based on satellitederived cloud data. The SEVIRI instrument, onboard the MSG satellites, is used to provide data regarding the cloud modification factor. These data are used, along with model simulations, performed with libradtran, to derive the global solar irradiance incident on a horizontal surface, a surface with a tilted orientation and the direct normal irradiance. The study focuses on the area of Greece and the work is part of the Hellenic Network for Solar Energy, developed to support solar energy applications. The daily amount of solar energy, as well as the monthly and annual sums, are estimated, during an 11-year period and a monthly climatology is derived. Results are compared with measurements from various ground stations in Greece. Comparison

12 12 shows a general good agreement between satellite and stations data, with the highest differences occurring in cases of broken cloud conditions or very thick clouds. Solar energy collected from surfaces under tilted orientations can provide % higher amounts than horizontal surfaces. In Greece, the highest collected monthly solar energy values are found during summer months, in Southern Peloponnese, Crete and the Cyclades islands, and exceed 250 kwh/m 2.

13 13 Εκτεταμένη περίληψη Τα αιωρούμενα σωματίδια είναι από τα πιο σημαντικά συστατικά της ατμόσφαιρας, καθώς επηρεάζουν την εισερχόμενη ηλιακή ακτινοβολία, αλλά και την υπέρυθρη ακτινοβολία που εκπέμπεται από τη Γη, μέσω διαδικασιών σκέδασης και απορρόφησης. Αυτή είναι γνωστή ως άμεση επίδραση των αιωρούμενων σωματιδίων. Η έμμεση επίδραση είναι αυτή, κατά την οποία τα αιωρούμενα σωματίδια επιφέρουν μεταβολές στις ιδιότητες των νεφών. Η παρουσία των αιωρούμενων σωματιδίων, έχει ως αποτέλεσμα τη μείωση του μεγέθους των υδροσταγονιδίων σε ένα νέφος και την αύξηση του αριθμού τους. Ως αποτέλεσμα, αυξάνεται η ανακλαστικότητα του νέφους, καθώς και η διάρκεια ζωής του, αφού τα μικρότερα σε μέγεθος υδροσταγονίδια, ελαττώνουν την ικανότητα υετού του νέφους. Η ημι-έμμεση επίδραση των αιωρούμενων σωματιδίων είναι αυτή, που προκαλεί μεταβολές στο ποσοστό κάλυψης των νεφών και στην ολική ποσότητα ύδατος σε υγρή μορφή, λόγω απορρόφησης της ακτινοβολίας από τα σωματίδια. Αυτές οι επιδράσεις έχουν ως αποτέλεσμα τα αιωρούμενα σωματίδια να παίζουν σημαντικό ρόλο στη μεταβολή του ενεργειακού ισοζυγίου του συστήματος Γη ατμόσφαιρα. Εκτός από τις επιδράσεις στο ενεργειακό ισοζύγιο, τα αιωρούμενα σωματίδια μπορεί να έχουν σοβαρές επιπτώσεις στην ανθρώπινη υγεία και την ατμοσφαιρική χημεία. Αν και βρίσκονται στο κέντρο της επιστημονικής έρευνας τα τελευταία χρόνια, η υψηλή χωρική και χρονική μεταβλητότητα τους και οι πολύπλοκες αλληλεπιδράσεις τους στην ατμόσφαιρα, προκαλούν δυσκολίες στον καθορισμό του κλιματικού τους εξαναγκασμού, που εξακολουθεί να έχει υψηλή αβεβαιότητα. Τα νέφη είναι ο κύριος ατμοσφαιρικός παράγοντας που επηρεάζει τη μεταβολή της ηλιακής ακτινοβολίας, σε ημερήσια βάση. Οι χρονικές μεταβολές των νεφών καθορίζουν τη διαθεσιμότητα της ηλιακής ενέργειας σε μια περιοχή. Τα νέφη ανακλούν την εισερχόμενη ηλιακή ακτινοβολία, οδηγώντας στην ψύξη του συστήματος Γη ατμόσφαιρα, ενώ απορροφούν την εκπεμπόμενη υπέρυθρη ακτινοβολία και συνεισφέρουν στη θέρμανση των θερμοκηπικών αερίων. Η επίδραση των νεφών στην ηλιακή ακτινοβολία εξαρτάται κυρίως από τον τύπο του νέφους, το μέγεθος των υδροσταγονιδίων ή παγοκρυστάλλων του και το ύψος στο οποίο βρίσκεται.

14 14 Στην παρούσα διατριβή, μελετούνται οι οπτικές ιδιότητες των αιωρούμενων σωματιδίων και των νεφών, με ιδιαίτερο ενδιαφέρον να δίνεται στην επίδραση, που έχουν αυτές οι ιδιότητες, στη διάδοση της ηλιακής ακτινοβολίας στην ατμόσφαιρα. Στο 1 ο κεφάλαιο, περιγράφονται οι βασικές αρχές της διάδοσης της ηλιακής ακτινοβολίας. Δίνονται οι νόμοι ακτινοβολίας μελανού σώματος και διάφορες παραδοχές για τη λύση της εξίσωσης διάδοσης της ακτινοβολίας. Περιγράφονται οι βασικές αρχές της θεωρίας της σκέδασης και δίνεται περιγραφή των βασικών χαρακτηριστικών της σκέδασης Rayleigh και της σκέδασης Mie. Ακολουθεί περιγραφή των κυρίων χαρακτηριστικών των αιωρούμενων σωματιδίων και νεφών, δίνοντας έμφαση στους τύπους τους και τις ιδιότητες που επηρεάζουν τη διάδοση της ακτινοβολίας. Το 2 ο κεφάλαιο παρέχει μια περιγραφή των βάσεων δεδομένων που χρησιμοποιήθηκαν, μαζί με περιγραφή του μοντέλου διάδοσης ακτινοβολίας. Περιγράφονται το επίγειο δίκτυο σταθμών AERONET, το πρόγραμμα σύγκρισης μοντέλων αιωρούμενων σωματιδίων AeroCom, το δορυφορικό όργανο MODIS και οι δορυφόροι MSG. Τέλος παρουσιάζονται τα βασικά στοιχεία του μοντέλου διάδοσης ακτινοβολίας uvspec, περιγράφοντας τις επιλογές δεδομένων εισόδου και αποτελεσμάτων που παρέχονται. Στο κεφάλαιο 3, οι οπτικές ιδιότητες των αιωρούμενων σωματιδίων, από τρεις διαφορετικές βάσεις δεδομένων, χρησιμοποιήθηκαν για τον υπολογισμό της υπεριώδους (UV) και ορατής (VIS) ηλιακής ακτινοβολίας, στην περιοχή της Μεσογείου. Οι οπτικές ιδιότητες των αιωρούμενων σωματιδίων, από την κλιματολογία AeroCom και το δορυφορικό όργανο MODIS, χρησιμοποιήθηκαν για τη μελέτη των διαφορών τους από μετρήσεις στο έδαφος, που παρέχονται από το δίκτυο AERONET. Υπολογίστηκε η επίδραση των διαφορετικών τιμών των οπτικών ιδιοτήτων, στον υπολογισμό της UV και VIS ακτινοβολίας, υπό ανέφελο ουρανό, για 8 σταθμούς του AERONET, στην περιοχή της Μεσογείου. Χρησιμοποιήθηκαν δεδομένα αιωρούμενων σωματιδίων επιπέδου 2.0 του AERONET, για τους 8 σταθμούς, για τη χρονική περίοδο Υπολογίστηκαν οι μέσες μηνιαίες τιμές των συντελεστών α και β του Ångstrom, της ανακλαστικότητας μεμονωμένης σκέδασης και του παράγοντα ασυμμετρίας, οι οποίες στη συνέχεια αποτέλεσαν δεδομένα εισόδου στο μοντέλο libradtran. Για τον υπολογισμό της ακτινοβολίας στο UV, οι α και β συντελεστές Ångstrom προέκυψαν από τα οπτικά βάθη στα 340 and 380 nm, ενώ η ανακλαστικότητα μεμονωμένης σκέδασης και ο παράγοντας ασυμμετρίας

15 15 επιλέχθησαν στο χαμηλότερο μήκος κύματος που παρέχεται, για αυτές τις παραμέτρους, από το AERONET, τα 440 nm. Αντίστοιχα για τον υπολογισμό της VIS ακτινοβολίας, οι συντελεστές Ångstrom υπολογίστηκαν από τα οπτικά βάθη στα 440 και 675nm, ενώ για την ανακλαστικότητα μεμονωμένης σκέδασης και τον παράγοντα ασυμμετρίας επιλέχθησαν οι μέσες τιμές σε αυτά τα μήκη κύματος. Αντίστοιχα για το AeroCom, οι συντελεστές Ångstrom προέκυψαν από τα οπτικά βάθη στα 345 και 380 nm (για το UV ) και 440 και 665 nm (για το VIS). Από τα δεδομένα του MODIS, χρησιμοποιήθηκαν οι μηνιαίες κλιματολογικές τιμές του συντελεστή Ångstrom και του οπτικού βάθους στα 550nm για την περίοδο Οι αντίστοιχες τιμές στο UV υπολογίστηκαν με τη φόρμουλα του Ångstrom τ(λ 1 )=τ(λ 2 )*(λ 1 /λ 2 ) -α, αφού δεν παρέχονται δεδομένα από το MODIS στο υπεριώδες. Για την ανακλαστικότητα μεμονωμένης σκέδασης και τον παράγοντα ασυμμετρίας χρησιμοποιήθηκαν οι σταθερές τιμές 0.95 και 0.67, αντίστοιχα. Οι μέσες μηνιαίες τιμές όλων αυτών των παραμέτρων, από κάθε βάση δεδομένων χωριστά, εισήχθησαν στο μοντέλο UVspec του λογισμικού πακέτου LibRadtran, για τον υπολογισμό της ακτινοβολίας στο UV και VIS. Οι υπολογισμοί πραγματοποιήθηκαν για την 15 η ημέρα του κάθε μήνα. Οι απόλυτες διαφορές στο οπτικό βάθος των αιωρούμενων σωματιδίων, μεταξύ AeroCom και AERONET, μπορεί να φτάσουν τα 0.25 και 0.15, στο UV και VIS κομμάτι του ηλεκτρομαγνητικού φάσματος, αντίστοιχα. Αυτές οι μέγιστες διαφορές παρατηρούνται την άνοιξη, όπου τα οπτικά βάθη είναι μεγαλύτερα. Η μέγιστη απόλυτη διαφορά, στο οπτικό βάθος, μεταξύ του AERONET και του MODIS, στα 550 nm, παρατηρείται επίσης την άνοιξη, αλλά είναι πολύ μικρότερη της αντίστοιχης μεταξύ AeroCom και AERONET, έχοντας την τιμή των Κατά τη διάρκεια του υπόλοιπου έτους, η κλιματολογία AERONET είναι σε καλύτερη συμφωνία με τα δεδομένα του MODIS, σε σχέση με αυτά που παρέχονται από το AeroCom. Οι μέσες μηνιαίες τιμές των οπτικών ιδιοτήτων των αιωρούμενων σωματιδίων, από το AERONET, το AeroCom και το MODIS, χρησιμοποιήθηκαν για τον υπολογισμό της UV ( nm) και VIS ( nm) ηλιακής ακτινοβολίας στο έδαφος. Οι διαφορές μεταξύ AERONET και AeroCom, στο UV, μπορεί να φτάσουν τα 7 (12%) και 4.5 W/m 2 (15%), για το τοπικό μεσημέρι και τη μέση τιμή της ημέρας,

16 16 αντίστοιχα. Στο VIS κομμάτι του φάσματος, οι αντίστοιχες διαφορές είναι 35 (10%) και 30 W/m 2 (12%). Οι μεγαλύτερες διαφορές παρατηρούνται στους σταθμούς της Βόρειας Αφρικής και της Μέσης Ανατολής, στα παράλια της Μεσογείου (Nes Ziona, Sede Boker, Blida) και στην Κρήτη. Οι διαφορές μεταξύ του AERONET και του MODIS, στο VIS, είναι χαμηλότερες, φτάνοντας τα 20 (6%) και 15 W/m 2 (7%), για το τοπικό μεσημέρι και την μέση τιμή της ημέρας, αντίστοιχα. Οι υψηλότερες διαφορές παρατηρούνται στο σταθμό της Κρήτης (FORTH) και στη Nes Ziona. Κατά μέσο όρο, οι διαφορές στο UV είναι μεταξύ 0 και 3%. Κατά τη διάρκεια του χειμώνα και της άνοιξης, οι ακτινοβολίες στο UV, που προκύπτουν από τα δεδομένα του MODIS, είναι χαμηλότερες κατά περίπου 3%, ενώ η μέση διαφορά ελαττώνεται γύρω στο 0% για το υπόλοιπο έτος. Οι υψηλότερες διαφορές παρατηρούνται στον σταθμό Blida (6%). Η αντίστοιχη διαφορά στην υπολογιζόμενη ακτινοβολία στο έδαφος, στο UV, είναι 2 και 3.3 W/m 2, για τη μέση τιμή της ημέρας και το τοποικό μεσημέρι, αντίστοιχα. Γενικά, οι διαφορές, που αναφέρθηκαν παραπάνω, στις υπολογιζόμενες ακτινοβολίες, οφείλονται κατά κύριο λόγο στις τιμές του οπτικού βάθους, ενώ η ανακλαστικότητα μεμονωμένης σκέδασης παίζει δευτερεύοντα ρόλο. Οι κλιματολογίες των οπτικών ιδιοτήτων των αιωρούμενων σωματιδίων, από το MODIS και το AeroCom, παρουσιάζουν τις μεγαλύτερες διαφορές, σε σχέση με τις αντίστοιχες τιμές του AERONET, σε περιοχές όπου η παρουσία αιωρούμενων σωματιδίων ερημικής σκόνης είναι κυρίαρχη. Οι τιμές του MODIS είναι σε καλύτερη συμφωνία με αυτές του AERONET, κάτι που υποδεικνύει την ανάγκη για βελτίωση του κύκλου ζωής της ερημικής σκόνης στα μοντέλα του AeroCom. Μια ολοκληρωμένη σύγκριση επιχειρήθηκε, μεταξύ κλιματολογιών των οπτικών ιδιοτήτων των αιωρούμενων σωματιδίων, στην περιοχή της Μεσογείου, από το AeroCom, το MODIS και το AERONET, καθώς και μεταξύ των αντίστοιχων υπολογιζόμενων ακτινοβολιών στο έδαφος, στο UV και VIS κομμάτι του ηλεκτρομαγνητικού φάσματος. Σύμφωνα με τα αποτελέσματα, οι βάσεις δεδομένων του δορυφόρου και των μοντέλων, χρειάζονται ακόμα βελτιώσεις και οι κλιματολογίες περαιτέρω μελέτες. Τα δεδομένα του MODIS είναι σε καλύτερη συμφωνία με τις επίγειες μετρήσεις και θα μπορούσαν να προτιμηθούν, για υπολογισμούς της UV και VIS ηλιακής ακτινοβολίας. Παρόλα αυτά, με επιπλέον βελτιώσεις, η κλιματολογία AeroCom μπορεί να είναι ένα σημαντικό εργαλείο, αφού παρέχει δεδομένα των οπτικών ιδιοτήτων των αιωρούμενων σωματιδίων, φυσικής και ανθρωπογενούς προέλευσης, με υψηλή φασματική ανάλυση.

17 17 Στο κεφάλαιο 4, υπολογίζονται η επίδραση των αιωρούμενων σωματιδίων στην ακτινοβολία και η ανακλαστικότητα μεμονωμένης σκέδασης, στο UV, σε έναν τυπικό σταθμό της Δυτικής Ευρώπης. Μετρήσεις του οπτικού βάθους, ολικού όζοντος και υπεριώδους ακτινοβολίας, παρέχονται από το όργανο Brewer#178, που βρίσκεται σε λειτουργία στην περιοχή Uccle του Βελγίου. Τα δεδομένα αυτά χρησιμοποιούνται για πρώτη φορά, για τον υπολογισμό της ικανότητας κλιματικού εξαναγκασμού (radiative forcing efficiency, RFE) των αιωρούμενων σωματιδίων, στο UV κομμάτι του ηλεκτρομαγνητικού φάσματος. Η μελέτη αφορά τη χρονική περίοδο Ιούλιος 2006 Μάιος Το οπτικό βάθος των αιωρούμενων σωματιδίων εξάγεται από τις μετρήσεις άμεσης ακτινοβολίας του Brewer, χρησιμοποιώντας τη μέθοδο Langley και παρέχονται σε 5 μήκη κύματος (306.3, 310.1, 313.5, και nm). Τα δεδομένα του οπτικού βάθους από το Brewer, στα nm, είναι σε πολύ καλή συμφωνία με αυτά που εξάγονται, στο ίδιο μήκος κύματος, από τα δεδομένα του οργάνου CIMEL, που βρίσκεται σε λειτουργία σε κοντινή απόσταση, ως μέρος του δικτύου AERONET. Το οπτικό βάθος του CIMEL στα nm εξάγεται από το αντίστοιχο στα 340 nm, χρησιμοποιώντας τη φόρμουλα του Ångstrom. Η σύγκριση του οπτικού βάθους των 2 οργάνων δίνει τιμές για την κλίση της ευθείας, το σημείο τομής με τον κατακόρυφο άξονα και το συντελεστή προσδιορισμού, ίσες με , και 0.947, αντίστοιχα. Το RFE των αιωρούμενων σωματιδίων υπολογίζεται στο κομμάτι nm και nm του ηλεκτρομαγνητικού φάσματος, παίρνοντας υπόψιν τις αντίστοιχες μετρήσεις του Brewer στο κάθε φασματικό κομμάτι και θεωρητικούς υπολογισμούς, υπό ανέφελο ουρανό και συνθήκες έλλειψης αιωρούμενων σωματιδίων, οι οποίες πραγματοποιούνται με το μοντέλο uvspec. Το RFE των αιωρούμενων σωματιδίων αυξάνει με τη μείωση της ηλιακής ζενίθιας γωνίας. Στο κομμάτι του φάσματος, nm, οι υψηλότερες τιμές παρατηρούνται στις 30 (-6.9 ± 0.9 W/m 2 ), ενώ στις 60 το αντίστοιχο RFE είναι 2.5 φορές μικρότερο (-2.7 ± 0.1 W/m 2 ). Στο UV-Β ( nm), η τιμή του RFE στις 60 ( ± W/m 2 ) είναι 5 φορές μικρότερη από την αντίστοιχη στις 30 (-0.35 ± 0.04 W/m 2 ). Μια μικρή μείωση παρατηρείται επίσης για το RFE (%) με την αύξηση της ζενίθιας γωνίας, που βρίσκεται μέσα στις αβεβαιότητες της μέτρησης.

18 18 Θεωρητικοί υπολογισμοί, με το μοντέλο διάδοσης ακτινοβολίας uvspec, προσθέτοντας δεδομένα για τα αιωρούμενα σωματίδια, χρησιμοποιήθηκαν, μαζί με τις μετρήσεις του Brewer στο UV, για τον υπολογισμό της ανακλαστικότητας μεμονωμένης σκέδασης σε 5 μήκη κύματος του UV (306.5, 310, 313.5, και 320 nm). Ένα Look-Up-Table (LUT) δημιουργήθηκε, παρέχοντας την ολική ακτινοβολία, σε κάθε μήκος κύματος, συναρτήσει της ηλιακής ζενίθιας γωνίας, του ολικού όζοντος, του οπτικού βάθους των αιωρούμενων σωματιδίων και της ανακλαστικότητας μεμονωμένης σκέδασης. Κάθε μέτρηση ακτινοβολίας συγκρίθηκε με την τιμή του μοντέλου, που αντιστοιχεί στην ζενίθια γωνία, το οπτικό βάθος και το όζον της μέτρησης. Μόνο οι τιμές ανακλαστικότητας μεμονωμένης σκέδασης, για τις οποίες οι διαφορές ανάμεσα στην μέτρηση και τη θεωρητική τιμή ήταν μικρότερες του 1%, έγιναν δεκτές. Ο αριθμός των αποδεκτών τιμών, με αυτόν τον τρόπο, εξαρτάται από το οπτικό βάθος των αιωρούμενων σωματιδίων. Όσο μικρότερο είναι, τόσο μεγαλύτερος ο αριθμός των τιμών που προκύπτει για την ανακλαστικότητα μεμονωμένης σκέδασης, επομένως ελαττώνεται η ακρίβεια προσδιορισμού της παραμέτρου. Η καινοτομία αυτής της μελέτης, βρίσκεται στο ότι, εκμεταλλευόμενοι τις μετρήσεις οπτικού βάθους σε αυτά τα μήκη κύματος, μελετάται η εγκυρότητα της μεθόδου σε χαμηλά UV-Α και σε UV-Β μήκη κύματος, για πρώτη φορά στις φασματικές μετρήσεις των Brewer. Οι υπολογιζόμενες μέσες μηνιαίες τιμές της ανακλαστικότητας μεμονωμένης σκέδασης, που προκύπτει από τις μετρήσεις του Brewer, στα 320 nm, είναι σε καλή συμφωνία (± 0.01) με τις αντίστοιχες τιμές του οργάνου CIMEL, που παρέχονται στα 440 nm. Λόγω αυξανόμενων αβεβαιοτήτων και επίδρασης της απορρόφησης του όζοντος, υψηλότερες διαφορές παρατηρούνται στο μήκος κύματος των nm. Για τα υπόλοιπα μήκη κύματος, μέσες διαφορές της τάξης του 0.03 παρατηρήθηκαν. Στο κεφάλαιο 5, παρουσιάζεται μελέτη της επίδρασης των αιωρούμενων σωματιδίων, υπό ανέφελο ουρανό, στην άμεση ηλιακή ακτινοβολία σε επιφάνεια κάθετη στη διεύθυνση της ακτινοβολίας (DNI). Η μελέτη πραγματοποιείται για την περιοχή της Ευρώπης. Η άμεση επίδραση των αιωρούμενων σωματιδίων στην ημερήσια ποσότητα ενέργειας που αντιστοιχεί στην DNI, υπολογίζεται για κάθε ημέρα, χρησιμοποιώντας δεδομένα από το όργανο MODIS, του δορυφόρου Terra της NASA και συμπληρωματικά δεδομένα για τα αιωρούμενα σωματίδια από το δίκτυο AERONET. Η χρονική περίοδος μελέτης ξεκινάει

19 19 από την αρχή λειτουργίας του MODIS (Μάρτιος 2000) μέχρι το τέλος του Η DNI υπολογίζεται χρησιμοποιώντας το μοντέλο διάδοσης ακτινοβολίας SBDART και παρέχεται με χωρική ανάλυση 1 x1, μαζί με την αντίστοιχη τιμή του κλιματικού εξαναγκασμού των αιωρούμενων σωματιδίων (RF) για κάθε pixel του δορυφόρου. Το οπτικό βάθος, όπως παρέχεται από το MODIS, θεωρείται σταθερό κατά τη διάρκεια της ημέρας. Για κάθε pixel του δορυφόρου και κάθε ημέρα της εξεταζόμενης περιόδου, υπολογίζεται από το μοντέλο η DNI ακτινοβολία, με βήμα 30 λεπτών, από την ανατολή ως τη δύση του Ηλίου, ως συνάρτηση της ηλιακής ζενίθιας γωνίας και των συντελεστών α και β του Ångstrom. Οι ακτινοβολίες διορθώνονται για την επίδραση της απόστασης Γης Ηλίου. Η ενέργεια, που αντιστοιχεί στην DNI, υπολογίζεται στη συνέχεια, ολοκληρώνοντας πάνω στις μεμονωμένες τιμές της ημέρας. Η DNI υπολογίζεται και για συνθήκες έλλειψης αιωρούμενων σωματιδίων, ώστε να υπολογιστεί η άμεση επίδραση τους. Κατά τη διάρκεια των 13 ετών που εξετάζονται, επιλέγονται μόνο οι μήνες, για τους οποίους υπάρχει διαθεσιμότητα δεδομένων κατά ποσοστό τουλάχιστον 40% (12 τουλάχιστον ημέρες το μήνα). Αυτό έχει ως αποτέλεσμα, να γίνουν δεκτά μόνο τα αποτελέσματα της περιόδου Μάιος Σεπτέμβριος, κατά τη διάρκεια κάθε έτους, καθώς οι υπόλοιποι μήνες, λόγω αυξημένης νέφωσης, δεν πληρούσαν το κριτήριο. Οι μηνιαίες κλιματικές τιμές του συντελεστή α του Ångstrom, όπως παρέχονται από 37 επιλεγμένους σταθμούς του AERONET, επεξεργάστηκαν με τη μέθοδο της κανονικής βέλτιστης παρεμβολής, ώστε να παρέχονται στη χωρική ανάλυση του MODIS (1 x1 ). Οι χαμηλότερες τιμές του συντελεστή α του Ångstrom, γύρω στα 0.6, παρατηρούνται την άνοιξη, στις ακτές της Μεσογείου, ενδεικτικό στοιχείο των αιωρούμενων σωματιδίων ερημικής σκόνης, που επηρεάζουν την περιοχή. Μόνιμες υψηλές τιμές, ενδεικτικές της παρουσίας σωματιδίων ανθρωπογενούς προέλευσης, παρατηρούνται σε περιοχές της Κεντρικής Ευρώπης, με πιο χαρακτηριστική αυτή της κοιλάδας του Πάδου, στην Ιταλία. Οι υπολογιζόμενες τιμές του RF (%), είναι υψηλές σε περιοχές που επηρεάζονται από την παρουσία σωματιδίων ερημικής σκόνης και από έντονες ανθρωπογενείς δραστηριότητες, όπως οι ακτές της Μεσογείου και η κοιλάδα του Πάδου. Η εξασθένηση στην DNI, που οφείλεται στα αιωρούμενα σωματίδια,

20 20 σε αυτές τις περιοχές, μπορεί να φτάσει τιμές γύρω στο 35% και 35-45%, αντίστοιχα, που αντιστοιχεί σε 4 και kwh/m 2 την ημέρα. Υψηλές τιμές (30 35%) παρατηρούνται τον Ιούνιο και τον Ιούλιο, στις Κάτω Χώρες, στη Γερμανία και την Πολωνία. Στις περισσότερες περιοχές, σε περιόδους χαμηλού φόρτου αιωρούμενων σωματιδίων, η εξασθένιση στη DNI κυμαίνεται γύρω στο 20%, που σημαίνει ότι 2 με 3 kwh/m 2 λιγότερες λαμβάνονται την ημέρα. Η υπολογιζόμενη DNI, φαίνεται να έχει αυξητικές τάσεις, κατά τη διάρκεια των τελευταίων ετών της εξεταζόμενης περιόδου, λόγω μείωσης του φόρτου αιωρούμενων σωματιδίων, που παρατηρείται σε πολλές περιοχές της Ευρώπης. Οι μεγαλύτερες ετήσιες αυξήσεις κυμαίνονται γύρω στο 6 με 12%, που αντιστοιχούν σε ποσό της τάξης του 0.5 με 1.25 kwh/m 2 περισσότερο την ημέρα. Αυτές οι τιμές παρατηρούνται κυρίως στην Κεντρική και Ανατολική Ευρώπη, τα Βαλκάνια και τη χερσόνησο της Ιταλίας. Η DNI υπολογίστηκε ακόμα, παίρνοντας υπόψιν την κλιματολογία AOD του MODIS, που προκύπτει από την περίοδο των 13 ετών. Η DNI, που υπολογίζεται με βάση την κλιματολογία AOD, είναι λίγο χαμηλότερη από αυτήν που υπολογίζεται από τα ημερήσια δεδομένα AOD του MODIS. Ανάλογα με την ημέρα και την περιοχή μελέτης, υποεκτιμήσεις της τάξεως του 0.6 kwh/m 2 την ημέρα, μπορούν να προκύψουν, οι οποίες πρέπει να λαμβάνονται υπόψιν στην περίπτωση που επιλεχθεί να χρησιμοποιηθεί η κλιματολογία για υπολογισμούς ακτινοβολίας. Στην μελέτη αυτή, εξετάστηκε η επίδραση των αιωρούμενων σωματιδίων στην ημερήσια ενέργεια που λαμβάνεται από την άμεση ακτινοβολία, όταν αυτή προσπίπτει σε επιφάνεια κάθετη στη διεύθυνση της εισερχόμενης ακτινοβολίας. Αυτή η παράμετρος είναι υψηλής σημασίας για τα συστήματα ηλιακής ενέργειας, όπου απαιτείται γνώση των χωρικών και χρονικών μεταβολών της λαμβάνουσας ακτινοβολίας DNI και όχι τόσο της ολικής ακτινοβολίας σε οριζόντιο επίπεδο, που είναι μια παράμετρος για την οποία παρέχονται πιο συχνά και πιο εύκολα πληροφορίες. Η συνεχής ανάγκη για ανανεώσιμη ενέργεια και η αύξηση των συστημάτων παραγωγής στις περισσότερες περιοχές της Ευρώπης, θέτουν την ανάγκη για επιπλέον μελέτες της μεταβλητότητας της ηλιακής ακτινοβολίας και κυρίως της άμεσης συνιστώσας της. Στο κεφάλαιο 6, η ηλιακή ακτινοβολία στο έδαφος υπολογίζεται, παίρνοντας υπόψιν δεδομένα για τα νέφη που εξάγονται από δορυφορικές μετρήσεις. Το Ελληνικό Δίκτυο Ηλιακής Ενέργειας έχει

21 21 αναπτυχθεί πρόσφατα, για την υποστήριξη των εφαρμογών ηλιακής ενέργειας, παρέχοντας αναλυτικές πληροφορίες για την ηλιακή ακτινοβολία, που προσπίπτει σε επιφάνειες διαφόρων προσανατολισμών. Δορυφορικά δεδομένα, από το όργανο SEVIRI των δορυφόρων Meteosat Second Generation (MSG), τα οποία συμπληρώνονται με δεδομένα από το όργανο MODIS, των δορυφόρων Terra και Aqua, χρησιμοποιήθηκαν για τον υπολογισμό της επίδρασης των νεφών στην ηλιακή ακτινοβολία. Η επίδραση αυτή εκφράζεται με την παράμετρο CMF (cloud modification factor). Το CMF λαμβάνει τιμές από 0 εώς 1, με το 0 να αντιστοιχεί σε συνθήκες πλήρης νέφωσης και το 1 σε συνθήκες ανέφελου ουρανού. Η παράμετρος CMF παρέχεται από τον MSG με χωρική ανάλυση 5 km και χρονική ανάλυση 15 λεπτών. Δεδομένα CMF επεξεργάστηκαν για την περίοδο , ενώ πραγματοποιήθηκαν θεωρητικοί υπολογισμοί με μοντέλο διάδοσης ακτινοβολίας, για τον υπολογισμό της ηλιακής ακτινοβολίας στο έδαφος. Οι υπολογισμοί ηλιακής ακτινοβολίας περιλαμβάνουν την ολική ακτινοβολία σε οριζόντια επιφάνεια (GHI), την ολική ακτινοβολία σε επιφάνεια υπό κλίση και την άμεση ακτινοβολία, σε επιφάνεια κάθετη στη διεύθυνση των ακτίνων (DNI). Η ακτινοβολία υπολογίζεται πολλαπλασιάζοντας τον CMF με την αντίστοιχη θεωρητική τιμή ακτινοβολίας, υπό ανέφελο ουρανό, που υπολογίζεται με τη libradtran. Το μοντέλο χρησιμοποιείται για τον υπολογισμό της άμεσης και διάχυτης συνιστώσας, ως συνάρτηση της ηλιακής ζενίθιας γωνίας και του οπτικού βάθους των αιωρούμενων σωματιδίων. Για το οπτικό βάθος χρησιμοποιούνται κλιματικά δεδομένα από το MODIS, υπολογιζόμενα για την περίοδο Έχοντας την τιμή της παραμέτρου CMF από τον δορυφόρο, η ακτινοβολία στο έδαφος υπολογίζεται πολλαπλασιάζοντας την με τη θεωρητική τιμή του μοντέλου, που αντιστοιχεί στη ζενίθια γωνία και οπτικό βάθος της μέτρησης. Για τον υπολογισμό της ακτινοβολίας στην κεκλιμένη επιφάνεια, θεωρούμε, για κάθε pixel, ότι η γωνία κλίσης είναι ίση με το γεωγραφικό πλάτος του τόπου. Η ακτινοβολία παρέχεται, για κάθε ημέρα της εξεταζόμενης περιόδου, με υψηλή χρονική ανάλυση (15 λεπτά) και υπολογίζεται ακόμα η ημερήσια ποσότητα ακτινοβολίας, ολοκληρώνοντας τις τιμές κατά τη διάρκεια της ημέρας. Υπολογίζονται τα μηνιαία κλιματικά δεδομένα της διαθέσιμης ακτινοβολίας, όπως προκύπτουν από τα 11 έτη που μελετούνται.

22 22 Τα αποτελέσματα συγκρίνονται με επίγειες μετρήσεις, από 12 σταθμούς στην Ελλάδα. Οι ημερήσιες ακτινοβολίες, για το έτος 2012, συγκρίνονται, μεταξύ του δορυφόρου και των μετρήσεων των πυρανομέτρων από τους 12 σταθμούς. Παρατηρείται μια πολύ καλή συμφωνία μεταξύ των δορυφορικών και των επίγειων δεδομένων, με την κλίση της ευθείας που προκύπτει, να κυμαίνεται από 0.96 (σταθμός Θεσσαλονίκης) εώς 1.08 (σταθμός Ορεστιάδας). Οι υψηλότερες διαφορές παρατηρούνται σε συνθήκες κατακερματισμένων νεφών (broken clouds) και σε περιπτώσεις παρουσίας νεφών με πολύ υψηλό οπτικό βάθος. Η σχέση που προκύπτει είναι γραμμική και δεν υπάρχει εξάρτηση από την ημέρα του έτους και ως εκ τούτου, από την ηλιακή ζενίθια γωνία. Οι επιφάνειες συλλογής ηλιακής ενέργειας έχουν τη μεγαλύτερη αποδοτικότητα, όταν τοποθετούνται υπό κλίση, καθώς μπορούν να παρέχουν 15 με 25% υψηλότερα ποσά. Η μηνιαία διαθεσιμότητα ηλιακής ενέργειας, στην Ελλάδα, αποκτά τις υψηλότερες τιμές της τους καλοκαιρινούς μήνες, στη Νότια Πελοπόννησο, την Κρήτη και τα νησιά των Κυκλάδων, όπου ξεπερνά τις 250 kwh/m 2. Η ετήσια ποσότητα της λαμβανόμενης ηλιακής ακτινοβολίας σε οριζόντια επιφάνεια, κυμαίνεται από kwh/m 2 στη Βόρεια Ελλάδα μέχρι kwh/m 2 στη Νότια Πελοπόννησο, την Κρήτη και τα νησιά. Όταν η επιφάνεια είναι τοποθετημένη υπό κλίση, οι τιμές αυξάνονται κατά περίπου 20% στη Βόρεια Ελλάδα και γύρω στο 10% στις νότιες περιοχές, όπου η διαθεσιμότητα είναι ήδη αρκετά υψηλή. Αυτό έχει ως αποτέλεσμα οι ετήσιες τιμές ενέργειας να υπερβαίνουν τις 2000 kwh/m 2 σε αυτές τις περιοχές. Το Ελληνικό Δίκτυο Ηλιακής Ενέργειας δίνει τη δυνατότητα, για πρώτη φορά, της μελέτης της χρονικής εξέλιξης της νέφωσης και της ηλιακής ακτινοβολίας στο έδαφος, με υψηλή χωρική και χρονική ανάλυση, ενώ παρέχονται πληροφορίες για την ηλιακή ενέργεια για τα τελευταία 11 έτη, μαζί με την υποστήριξη επίγειων μετρήσεων. Τέλος, στο κεφάλαιο 7, γίνεται μια ανασκόπηση των βασικών αποτελεσμάτων και συμπερασμάτων, που προέκυψαν κατά τις μελέτες των προηγούμενων κεφαλαίων.

23 Table of Contents List of Acronyms Radiative Transfer in the Atmosphere Sun radiation Basic radiometric quantities Blackbody Radiation Planck law Stefan-Boltzmann law Wien law Kirchoff law Radiative transfer Beer-Bouguer-Lambert law Schwarzschild equation Plane-parallel atmosphere Scattering in the Atmosphere Rayleigh scattering Mie scattering Aerosols Aerosol Types Aerosol Size Distribution Radiative Effects Clouds Cloud Types Radiative Effects References Data sources from remote sensing instruments and models The AERONET network The AeroCom Project Meteosat Second Generation (MSG) satellites MODIS satellite instrument The Radiative Tranfer Model About uvspec Spectral Calculations and Radiative Transfer Solvers Atmospheric input data References Aerosol optical properties and corresponding solar irradiances in the Mediterranean

24 Abstract Introduction Data and Methodology Results and Discussion Conclusions References Aerosol optical properties and radiative effects in the UV Abstract Introduction Data and Location Methodology Results Comparison with the CIMEL AOD Aerosol Radiative Forcing Single Scattering Albedo Retrieval Conclusions References Aerosol effects on Direct Normal Irradiance in Europe Abstract Introduction Data and Methodology Results The Ångström α exponent seasonal variability DNI changes due to the AOD variability DNI derived from MODIS AOD climatology Conclusions References Solar irradiance calculations from satellite-derived cloud properties Abstract Introduction Data and Methodology Results Satellite-derived CMF compared with ground-measurements Retrieval of Solar Irradiance Conclusions References

25 7. Conclusions Acknowledgements

26 26 List of Acronyms AeroCom AERONET AIRS AMSR-E AMSU AOD ARM ASTER ATSR AVHHR BSRN CALIPSO CARL CERES CMF CRF DIAL DLR DMS DNE DNI DOE EOS ERBE ERBS ERS EUMETSAT FORTH GCOM-W1 GEO GERB GHI GOCART HITRAN HNSE HRV IMS-METU IPCC IR Aerosol Comparisons between Observations and Models Aerosol Robotic Network Atmospheric Infrared Sounder Advanced Microwave Scanning Radiometer - Earth Observing System Advanced Microwave Sounding Unit aerosol optical depth Atmospheric Radiation Measurement Advanced Spaceborne Thermal Emission and Reflection Radiometer Along Track Scanning Radiometer Advanced Very High Resolution Radiometer Baseline Surface Radiation Network Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations CART (Cloud and Radiation Testbed) Raman Lidar Clouds and Earth's Radiant Energy System cloud modification factor Climate Research Facility Differential Absorption Lidar German Aerospace Center dimethyl sulfide direct normal energy direct normal irradiance Department of Energy Earth Observing System Earth Radiation Budget Experiment Earth Radiation Budget Satellite European Remote Sensing Satellite European Organization for the Exploitation of Meteorological Satellites Foundation for Research and Technology Global Change Observation Mission 1st - Water geostationary orbit Geostationary Earth Radiation Budget global horizontal irradiance Global Ozone Chemistry Aerosol Radiation and Transport High Resolution Transmission Hellenic Network of Solar Energy high-resolution visible channel Middle East Technical University - Institute of Marine Sciences Intergovernmental Panel on Climate Change infrared

27 27 ISCCP LEO LOWTRAN LUT LW LWIR MCP MISR MODIS MOPITT MSG MWIR NASA NILU NIR NOAA OMI PM POLDER PSC RF RFE S&R SBDART SeaWiFFS SEVIRI SSA SW SWIR TOA TOMS TRACE-P UTC UV VIS VOC WMO International Satellite Cloud Climatology Project low earth orbit Low Resolution Transmission look-up-table long-wave long-wave infrared Mission Communication Payload Multi-angle Imaging Spectroradiometer Moderate Resolution Imaging Spectroradiometer Measurements of Pollution in the Troposphere Meteosat Second Generation mid-wave infrared National Aeronautics and Space Administration Norwegian Institute for Air Research near-infrared National Oceanic and Atmospheric Administration Ozone Monitoring Instrument particle mass Polarization and Directionality of Earth's Reflectances polar stratospheric clouds radiative forcing radiative forcing efficiency Search and Rescue Santa Barbara DISORT Atmospheric Radiative Transfer Sea-viewing Wide Field-of-view Sensor Spinning Enhanced Visible and Infrared Imager single scattering albedo short-wave short-wave infrared top of the atmosphere Total Ozone Mapping Spectrometer Transport and Chemical Evolution over the Pacific Coordinated Universal Time ultraviolet visible volatile organic compounds World Meteorological Organization

28 28 1. Radiative Transfer in the Atmosphere 1.1 Sun radiation The Sun s radiation is the main source and regulator of energy distribution in the Earth-atmosphere system, leading to the sustainability of life and chemical interactions between the various constituents of the atmosphere. The Sun is a dwarf star in the main sequence, aging around 4.6 billion years. Its emitted radiation undergoes several variations, the most important of which is the 11-year cycle, which describes the change in the activity of the Sun, accompanied with change in the number of sunspots on the surface. A slower, in time frequency, cycle is the variation known as the Milankovic cycles, which include changes in the eccentricity of the Earth s orbit around the Sun, with a cycle of around 100,000 years, changes in the tilt of the Earth s axis in relation to the orbital plane, with a cycle of around 41,000 years and finally, changes in the axial precession, with a cycle of around 26,000 years. From the time the Sun joined the main sequence, its luminosity has increased by about 30%. The emitted radiation intensity from the Sun has the Planck distribution of a blackbody radiation, corresponding to an effective temperature of around 5800 K, which is the temperature of the Sun s photosphere. The total emitted solar irradiance at the average Sun-Earth distance is called the solar constant and is equal to 1366 W m -2. The solar spectrum is presented in Figure 1.1.

29 29 Figure 1.1: Spectrum of a blackbody radiation (T= 6000 K) (dashed line) and of the irradiance at the top of the atmosphere (black solid line) and at sea level (gray shaded area). Source: Andrews (2010). The dashed line corresponds to the theoretical Planck distribution of radiation of a blackbody with effective temperature of 6000 K. The solar spectrum, as it reaches the top of the atmosphere, is presented with the black solid line and differs from the theoretical one, due to the slightly lower Sun s temperature, but mostly because of absorption and emission processes that take place as the radiation travels through the Sun s atmosphere. The solar radiation, that reaches the top of the atmosphere, is further diminished by the atmospheric constituents as it traverses the Earth s atmosphere and as a result the incident solar radiation at the sea level is represented by the gray shaded area on the graph. The basic atmospheric gases are also shown in the figure, in the area of the spectrum where they present their most efficient extinction on radiation. Ozone (O 3 ) shows large extinction efficiency in the ultraviolet and visible part of the spectrum,

30 along with oxygen (O 2 ), while water vapor (H 2 O) and carbon dioxide (CO 2 ) are the main absorbing gases in the infrared wavelengths Basic radiometric quantities The theory of radiative transfer is described in details in the work, amongst others, of Liou (2002), Vardavas and Taylor (2007) and Salby (2012). A brief description, based on these works, is provided in the next paragraphs. Considering a sphere of area A, the radiation emitted from an element of area da, at its centre, is considered to be transferred through an element of solid angle dω. The solid angle is defined by a cone with a base dσ at the surface of the sphere and the top of its vertex at the centre of the sphere (Figure 1.2). The element solid angle is given by d d (1.1) 2 r which in polar coordinates is written as ( rd )( r sind) d sindd (1.2) 2 r where θ, φ are the zenith and azimuth angles, respectively and r is the radius of the sphere.

31 31 Figure 1.2: Demonstration of a differential solid angle in polar coordinates. Source: Liou (2002). The amount of radiant energy de λ, that is emitted or passes through an area da in a time interval of dt, with a wavelength range of dλ, is given by de I cos dad ddt (1.3) This equation provides the monochromatic intensity or radiance, I λ, I de (1.4) cos dad ddt which describes the radiation in terms of energy per unit of area, steradian, wavelength and time. By integrating over all solid angles of the hemisphere, the parameter that is derived is the monochromatic flux or monochromatic irradiance, F λ, describing the radiative flux per unit of area, usually provided in Watts per m 2, I cos d (1.5) F

32 32 which in polar coordinates is consequently written as: 2 / 2 I 0 0 F (, )cos sindd. (1.6) Considering the radiation to be isotropic, that is the radiance does not depend on the direction, the monochromatic irradiance is then reduced to F I (1.7) where π has units of steradians. The total flux density or total irradiance, F, is derived by integrating over all wavelengths: 0 F F d. (1.8) 1.3. Blackbody Radiation Planck law A blackbody is a surface that is considered to be a perfect emitter and absorber. It is visualized with a cavity having a small opening and being in thermal equilibrium. All the radiation that enters the cavity is reflected upon the walls until finally being absorbed by it, therefore the inside of the cavity is perceived as black. The radiation emitted by the walls of the cavity undergoes continuous absorption and reflections, until absorption and emission reach an equilibrium. The radiation emitted by a blackbody depends on the temperature of the surface and has the maximum possible values at each wavelength, that correspond to that temperature. The radiation is characterized by isotropic distribution. The ideal blackbody emits radiance at a given frequency ν, given by: B 3 2h T) h e ( / kt 2 / c 1 (1.9)

33 33 where T is the surface s temperature, c is the speed of light in a vacuum, h = 6.626x10-34 J sec and k = x10-23 J deg -1 are the Planck s and Boltzmann s constant, respectively. In terms of wavelength, the blackbody radiance is written as: B 2hc T) hc / k e 5 / C C 1 ( e 5 / 2 1 ( T / T 2 1) (1.10) where 2 C hc, C hc/ k are the first and second radiation constants respectively. The above equations are known as the Planck s law. Figure 1.3: The spectrum of blackbody radiation, for an effective temperature of 200 and 300 K. Source: Vardavas and Taylor (2007). Figure 1.3 presents the emitted radiance for two blackbodies with temperatures of 300 and 200 K, respectively. It s observed that as the blackbody s temperature decreases, the maximum of its emitted energy is moved towards higher wavelengths. The radiation curve of the higher temperature s body is always above the lower temperature s one and is also more steep, meaning that the rise to the maximum emission wavelength is more rapid. The radiation that is emitted from Sun and Earth can be considered having the

34 distribution of the radiation emitted from a blackbody with a surface temperature of 5780 K and 288 K, respectively Stefan-Boltzmann law In order to calculate the total radiance emitted by a blackbody, the equation (1.10) is integrated over the entire wavelength spectrum 2 5 2hc / B( T) B ( T) d 0 d (1.11) 0 hc / kt e 1 Setting the variable x as, x hc / kt and since blackbody radiation is considered isotropic, the total flux is given by F 2k T x dx. (1.12) h c e ( T) B( T) x k Defining the parameter σ as,, the total blackbody flux is given by h c F( T) T 4. (1.13) The parameter σ is known as the Stefan-Boltzmann constant and is equal to 5.67x10-8 J m -2 K -4 s -1, while the derived equation (1.13) is defined as the Stefan-Boltzmann law Wien law The wavelength, at which the maximum emitted radiance by the blackbody is observed, is inversely proportional to the surface s temperature. This is known as the Wien s displacement law. In order to derive this wavelength, the derivative of the blackbody s radiance with regards to wavelength is set to zero

35 B ( T ) 0 35 (1.14) 2897 and the wavelength of maximum emission is given by max ( m). (1.15) T As the temperature of the blackbody increases, the emitted maximum energy is moved towards lower wavelengths, as already seen in Figure Kirchoff law A body s state of thermodynamic equilibrium means that the body emits and absorbs radiation at the same wavelength, while the amount of energy depends on its temperature. This concept is described by Kirchoff s law A, (1.16) where ε λ is the body s emissivity at wavelength λ, defined as the ratio of the intensity that is emitted to Planck s function, B λ (Τ) and Α λ is the body s absorptivity at the same wavelength, defined as the ratio of the intensity that is absorbed to B λ (Τ). The blackbody s emitted and absorbed radiation is the maximum possible, as mentioned before, therefore the equation (1.16) for a blackbody becomes: A 1. (1.17) A body which presents incomplete absorption and emission is called a gray body and in this case the equation (1.16) takes the form: A 1. (1.18)

36 Radiative transfer When radiation of intensity I λ passes through a medium of thickness equal to ds, then its intensity will be modified by an amount of di λ, equal to di k I ds, (1.19) where ρ is the medium s density and k λ is the mass extinction cross section. The extinction of the radiation as it travels through the medium is due to absorption and/or scattering, therefore the mass extinction cross section is the sum of two parts, the mass absorption and the scattering cross section. The intensity of the radiation, besides being reduced as it passes through the medium, it may also be enhanced by emission from the material s surface or radiation scattered in the direction of its propagation. Therefore the increase in the radiation s intensity is given by di j ds, (1.20) where j ν is the source function coefficient. As a result, the overall variation in the intensity will be the sum of reduction and increase that take place as it passes through the medium, di k I ds j ds. (1.21) With the definition of the source function, J λ, as the ratio of the source function coefficient to the mass extinction cross section, J j / k, (1.22) the equation (1.21) is written as di I k ds J (1.23) and this is the general form of the radiative transfer equation. The radiative transfer equation is very complex to be solved as it is, taking account the absorption, emission and multiple scattering processes that

37 are performed with the interaction of the radiation with the atmospheric constituents and usually several assumptions and approximations are made, depending on the nature of the problem that is studied Beer-Bouguer-Lambert law If the scattering and emissions processes are considered negligible, then, only absorption is responsible for the inflicted changes on the intensity of the radiation and the equation (1.23) becomes di I. (1.24) k ds If the radiation at travelling distance s=0, through the medium, is I λ (0), then the intensity at a distance s 1 is given by: 1 I ( s ) I (0) exp( k ds). (1.25) 1 0 s By assuming that the medium, through which the radiation travels, is homogeneous, then k λ can be taken out of the integral and after defining the path length, u, as s u 1 0 ds (1.26) the equation (1.25) takes the form: I ku ( s 1 ) I (0) e, (1.27) which is known as the Beer-Bouguer-Lambert law, stating that the decrease in the radiation intensity, as it travels through a homogeneous medium is exponential, depending on the medium mass extinction cross section and the path length traveled. Based on the Beer-Bouguet-Lambert law, the monochromatic transmissivity, T λ, is written as T I ) ku ( s 1 )/ I (0 e, (1.28)

38 38 whereas if no scattering is considered, the monochromatic absorptivity, A λ, is: A 1 ku 1 T e. (1.29) Schwarzschild equation If the medium considered is in local thermodynamic equilibrium and no scattering processes take place, then the only effects on the radiation are due to emission and absorption procedures. These assumptions are appropriate when considering the thermal infrared radiation emitted by the surface of the Earth. The source function in this case is equal to the Planck s function, J B (T) (1.30) and the radiative transfer equation takes the form: di I k ds B (T) (1.31) where the attenuation of the radiation s intensity is due to absorption as it travels through the medium and its enhancement is due to the blackbody emission from it. This is known as the Schwarzschild equation. Considering two distances traveled in the medium, s and s 1 (s < s 1 ), the monochromatic optical thickness of the layer that they enclose, is defined as s ( s, ) 1 1 s k ds. s (1.32) Since d ( 1 s, s) k ds (1.33) di ( s). (1.34) d ( s, s) equation (1.32) becomes I ( s) B T ( s) 1 By multiplying with ( s 1, s) e and consequently integrating from 0 to s 1, equation (1.34) becomes s ( s1, s) I ( s) e B T ( s) s 1 1 ( s1, s) d e 0 0 d ( s, s) and (1.35) 1

39 I s ( s1, s T ( s e k ds ( s,0) 1 1 ) ( s1) I (0) e B ). (1.36) 0 39 The radiation intensity I λ at point s 1, is attenuated by absorption from the medium, described by the first term on the right of equation (1.36) and enhanced by the radiation emitted from the layer between 0 and s 1, which in turn is described by the second part on the right Plane-parallel atmosphere In order to account for realistic procedures that take place into the atmosphere, one must consider the scattering along with absorption and emission processes. Adding the scattering parameters, into the radiative transfer equation, intensifies the complication and therefore several assumptions are made to derive solutions that target specific problems in a simpler manner, without lacking much in accuracy. To account for multiple scattering in the atmosphere, the scenario of plane-parallel atmosphere is usually used. The atmosphere is considered to be divided in parallel levels, in which the horizontal variability is considered negligible. That means that each level is considered to be horizontally homogeneous and the radiation field in the horizontal direction is isotropic. Plane-parallel atmosphere scenario doesn t take into account the sphericity of the Earth. As a result, the radiation intensity and all the atmospheric parameters present only vertical variability and equation (1.23) takes the form di( z;, ) cos kdz I( z;, ) J ( z;, ). (1.37) By defining the normal optical thickness, τ, as z kdz (1.38) the equation (1.37) for plane parallel atmospheres is written as di ( ;, ) I( ;, ) J ( ;, ) (1.39) d

40 40 where μ is the cosine of the zenith angle, μ=cosθ. In order to solve the radiative transfer equation for plane parallel atmospheres, the same process, as with the Schwarzschild equation, is followed. Figure 1.4: Demonstration of a plane-parallel atmosphere. Source: Liou (2002). The atmosphere is considered to have a top level, characterized by an optical thickness of τ=0, while the bottom level s optical thickness is defined as τ * (Figure 1.4). At an intermediate level, of optical thickness τ, the upward radiation intensity is I(τ; μ, φ), where μ>0 and the downward respectively is Ι(τ μ, φ), where μ<0. In order to derive the upward intensity at level τ, the equation (1.39) is multiplied with integrated from τ to τ *, / e and ( ) / * * ( ) / d I ( ;, ) I( *;, ) e J( ;, ) e (1 0) (1.40) while for the calculation of the downward intensity at level τ, the equation (1.39) is multiplied with and integrated from τ=0 to τ, / e

41 41 0) (1 ), ; ( ), (0; ), ; ( ) / ( 0 ) / ( d e J e I I. (1.41) In equations (1.40) and (1.41) the parameters I(τ * ; μ, φ) and I(0; -μ, φ) describe the radiation intensity incident in the atmosphere from the bottom and the top level, respectively. In order to derive the outward intensities at the bottom and top level, which are important parameters for many atmospheric applications, the optical thickness in equations (1.40) and (1.41) is set to 0 and τ *, respectively, providing, d e J e I I / 0 / * * * ), ; ( ), ; ( ), 0; ( (1.42) and d e J e I I ) / ( 0 / * * * * ), ; ( ), (0; ), ; (. (1.43) 1.5. Scattering in the Atmosphere The radiation, as it passes through a medium, interacts with matter in three different ways, through absorption, emission and scattering. A particle may absorb radiation and find itself in an excited energetic state or in a reverse way, it may emit radiation by falling into a state of lower energy. Scattering is the process by which a particle absorbs and consequently re-emits radiation, changing its frequency and/or direction, depending on the type of scattering. Detailed description of the interaction of matter and radiation, through scattering processes, can be found, amongst others, in Thomas and Stamnes (1999), Seinfeld and Pandis (2006) and Salby (2012). The basic concepts are described below. The extinction cross section defines the combined effect of both scattering and absorption, has units of area and is given by: abs scat ext. (1.44)

42 The corresponding extinction efficiency is defined by the ration of the extinction cross section, σ ext, to the cross-sectional area of the particle, A and is in turn the sum of the scattering and absorption efficiencies, 42 Q ext Q Q. (1.45) scat abs The ratio of a particle s scattering efficiency to the total extinction efficiency is defined as the single scattering albedo, Q scat scat. (1.46) Q ext C C ext The scattering of radiation by a particle includes three different processes which are: the elastic scattering, where radiation is re-emitted without any change in the wavelength, the quasi-elastic scattering, where the radiation is re-emitted with a shift in the wavelength due to Doppler and diffusion broadening and the inelastic scattering, where the radiation is re-emitted with a different wavelength. In the study of the atmospheric interaction of radiation with air molecules and aerosol particles, the scattering is considered to be elastic. The problem of the absorption and elastic scattering of radiation by a spherical particle is addressed in the Mie Debye Lorenz theory. The parameters that control the scattering process are the particle complex refractive index, the wavelength of the radiation and the size of the particle. The last two are combined in the size parameter α, which determines the type of elastic scattering that takes place, D p. (1.47) The scattering phase function defines the angular distribution of the scattered radiation and is defined as the scattered radiation intensity at angle θ, F(θ,α,m), normalized by the integral of the intensity at all possible angles, F(,, m) P(,, m), (1.48) F(,, m)sind 0

43 where m is the complex refractive index of the particle. Integration of the phase function over the entire sphere yields, P (,, m)sindd 4. (1.49) In the following,the phase function will be written as a function of only the angle θ, for simplicity purposes. The radiation intensity-weighted average of the cosine of the angle of scattering, provides the asymmetry factor, 1 cosf ( )sind 0 1 g cosp( )sind. (1.50) 2 0 F( )sind 2 0 The asymmetry factor describes the percent of the radiation scattered in the forward direction. A value of g=1 means that the radiation is completely scattered in the forward direction (θ=0 ), g=-1 corresponds to complete backscatter radiation (θ=180 ) and g=0 corresponds to isotropical scattering. Depending on the value of α, the scattering can be divided into three categories: α<<1, the Rayleigh scattering, where the particle size is very small compared to the wavelength of the incident radiation α 1, the Mie scattering, where the particle size is of the same magnitude as the radiation wavelength α>>1, the geometric scattering, where the particle size is very large compared to the wavelength of the radiation.

44 44 Figure 1.5: Radiation scattering from a spherical particle with a radius of 10-4 μm (a), 0.1 μm (b) and 1 μm (c). Source: Liou (2002). Figure 1.5 presents the scattering of radiation of wavelength λ=550nm, by a particle with a radius of (a) 10-4 μm, (b) 0.1 μm and (c) 1 μm. Small particles present isotropic scattering, with the scattered radiation having almost the same intensity in every direction. As the particle radius increases, the scattered radiation is no longer isotropic and the forward scattering is favored. For high radius values, the forward scattering becomes very large and the backward scattering is negligible.

45 45 Figure 1.6: The scattering efficiency as a function of the size parameter, for several values of the imaginary refractive index. Source: Liou (2002). Figure 1.6 depicts the scattering efficiency of a spherical scatterer, as a function of the size parameter. The scattering efficiency is plotted for several values of the imaginary component of the refractive index, m i, while the real part is considered to be equal to 1.5. For m i =0, which means that there is no absorption taking place, the scattering efficiency shows a rapid increase with the size parameter, reaching a maximum for λ α. For higher values of the size parameter, the scattering efficiency presents a series of maximum and minimum values, due to interference from light travelling though the sphere and being diffracted. As the size parameter gets even higher, the scattering efficiency reaches a value of 2 and presents a very small variation with wavelength. That means, that the scattering of radiation from very large particles is not a function of wavelength, a phenomenon that explains the white color of clouds. In cases where m i has positive values, so absorption processes take place, the behavior is similar with the scattering efficiency showing a decrease as expected. For high values of m i, the series of maximum and minimum values, that were observed for high values of the size parameter, disappear and as α obtains very high values, the scattering efficiency reaches values lower than 2, due to the increased attenuation that the radiation undergoes inside the sphere of the particle.

46 Rayleigh scattering In the Rayleigh scattering, radiation is scattered by particles, very small compared to the wavelength, as in the case of atmospheric molecules. The phase function and scattering cross section of the Rayleigh scattering are: 3 P ( ) (1 cos 2 ) (1.51) R( ) ( ) ap, (1.52) 3 where α P is the polarizability of the scatterer. Considering the complex refractive index, m, which is the sum of a real and imaginary part, m r and m i, describing the scattering and absorbing ability respectively, the polarizability of a group of scatterers is written as 2 3 m 1 a p ( ), (1.53) 2 4 n m 2 where n is the number density of the scattering medium. For wavelengths in the visible part of the spectrum, the absorption and by consequence m i is very small, whereas m r is around 1 and as a result the polarizability is written a p mr 1 (1.54) 2n and the scattering cross section of a single scatterer becomes ( m 1) ( ) r R. (1.55) 2 4 3n The Rayleigh scattering, as can be seen, is a function of λ -4, which means that as the wavelength decreases the scattering efficiency becomes much more intense. A prominent example is the blue colour of the skies, as a result of the much more efficient scattering of blue light wavelengths by the atmospheric molecules. During sunset, as the atmospheric path increases, the scattering of lower wavelengths is so high, that the

47 47 blue light wavelengths are completely scattered out of the observed direct beam and the sky at that directions acquires shades of red Mie scattering When the size of the particle is comparable or larger than the wavelength of the incident radiation, then the problem of the atmospheric scattering is addressed with the Mie scattering equations. The scattering, σ n and extinction, k n, cross section, for a single scatterer respectively are given by the following equations, ) 1)( (2 2 n n n n b a n k (1.56) 1 2 ) 1)Re( (2 2 n n n n b a n k k, (1.57) where k is the wavenumber, k=2π /λ. The coefficients α n and b n are given by ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( mx x m x mx mx x m x mx b mx x x mx m mx x x mx m a n n n n n n n n n n n n n n n n n n (1.58) where ψ n, ξ n are the Ricatti Bessel functions. The scattering phase function is ) ( 2 1 ) ( S 1 S P, (1.59) where S 1 and S 2 are the scattering amplitudes, given by ) ( 1) ( 1) (2 ) ( 1) ( 1) (2 n n n n n n n n n n b a n n n S b a n n n S (1.60) where

48 48 1 Pn ( ) n ( ) sin 1 dpn ( ) n ( ) d, (1.61) with P 1 ( ) being the associated Legendre polynomial. n 1.6. Aerosols In this paragraph, a brief description, based on the work by Seinfeld and Pandis (2006) and scientific reports from the Intergovernmental Panel on Climate Change (IPCC), on the aerosol types, sources, radiative effects etc., is provided. The liquid and solid particles, suspended in the atmosphere, due to various natural and anthropogenic processes, are known as atmospheric aerosols. Aerosols vary in shapes and sizes, chemical composition, emission sources and optical properties. The impact that they impose on living organisms and Earthatmosphere interactions, has mainly a local character, although their transport in large distances away from the emission sources is in some cases of extreme importance. Aerosols are emitted into the atmosphere from various natural mechanisms (windblown dust, sea spray, etc.) and anthropogenic activities (combustion processes, transportation, etc.). They are either emitted directly (primary aerosols) or formed in the atmosphere from precursor gases (secondary aerosols). Their diameter can vary from a few nanometers to several micrometers. Aerosols present high spatial distribution and small lifetimes, making their study and full understanding of their effects and interactions, a challenging task in present day scientific community. Figure 1.7 presents the various types of emission and removal processes, depending on the aerosol size. The distribution of aerosol sizes presents several modes. In general, aerosols are divided into fine and coarse particles. Fine particles have diameters less than 1 μm, while coarse particles are the largest aerosols with diameters higher than 1 μm. The fine particles present two different modes, the Aitken mode, with diameters lower than 0.1 μm and the accumulation mode with diameters between 0.1 and 1 μm. Fine particles are emitted through combustion processes or are formed in the atmosphere from chemical processes taking

49 49 place in precursor gases. These small particles subsequently take part in coagulation procedures and are removed from the atmosphere through rainout and washout processes, since they are not heavy enough to be removed with gravitational forces. Coarse mode particles are emitted in the atmosphere with mechanical processes (e.g. windblown dust) and dry deposition is the most effective removal process, due to their larger size. Figure 1.7: Size distribution of aerosols, sources and removal mechanisms. Source: Whitby and Cantrell (1976).

50 Aerosol Types The main atmospheric aerosol types, observed in the atmosphere, are particles of mineral dust, sea salt, sulfates and carbonaceous particles. Table 1.1 presents the global emission estimates for the basic aerosol types, emitted through natural or anthropogenic processes. Mineral dust particles are emitted in the atmosphere through mechanical processes, initiated from wind on arid and semiarid areas. Global emissions, as given by Zender et al. (2003) (Table 1.1) range from 48 to 609 Tg/yr, depending on the particle diameter, with the total number being around 1490 Tg/yr. The emission process of mineral dust depends on the type of surface and the wind velocity. The largest sources of mineral dust particles are the desert areas of Africa and Middle East. Dust particles live long enough in the atmosphere to be transported into large distances away from their source, when the atmospheric conditions are favorable. Dust particles emitted from the Sahara desert are often transported north, to Mediterranean cities, but are also observed to travel across the Atlantic Ocean and reach Eastern America territories. Mineral dust particles can have anthropogenic origins as well, due to land-use processes. Dust particles are amongst the largest of atmospheric aerosols and are mainly removed through dry deposition processes, although wet removal also takes place. Their radiative effect includes scattering and absorption of the incident solar radiation. Sea salt particles are the largest in size atmospheric aerosols, along with mineral dust. They are emitted from oceans as the wind interacts with the surface and from the evaporation of sea-spray during the break of waves. Gong et al. (2002) (Table 1.1) give a global emission estimate of Tg/yr. The most efficient removal of sea salt particles is achieved through dry deposition. As it is presented in Table 1.1, sea salt and mineral dust aerosols are responsible for the majority of aerosol mass in the atmosphere, even though the number of particles from these two types is lower than the number emitted from anthropogenic activities, which are mainly responsible for the production of smaller size particles. Sea salt particles are very efficient scatterers of solar radiation.

51 51 Carbonaceous aerosols include two main types of aerosols, black carbon and organic aerosol. Black carbon particles are primary aerosols, emitted through combustion processes, whereas organic matter can be emitted as primary aerosols or can be formed in the atmosphere from precursor gases. Global emission estimates of anthropogenic black carbon and organic aerosol, as given by Liousse et al. (1996), are around 12 and 81 Tg of carbon per year, respectively. Natural organic aerosols, formed from biogenic volatile organic compounds (VOC), have a burden of 11.2 Tg/yr (Chung and Seinfeld, 2002) (Table 1.1). Organic aerosols scatter solar radiation, while black carbon presents a significant absorption, leading to a heating of the atmosphere. Sulfate particles are secondary aerosols, formed in the atmosphere from gaseous precursors. They can be of natural origin, formed from sulfur dioxide (SO 2 ) which is emitted in the atmosphere during volcanic eruptions or formed from dimethyl sulfide (DMS) from marine plankton. The anthropogenic source of sulfate aerosols is through the emitted SO 2 in the atmosphere from fossil fuel burning. Global emission estimates for anthropogenic sulfates are around 48.6 Tg of sulfur per year (Liao et al., 2003), whereas natural sources burden the atmosphere with 20 Tg/yr (due to volcanic activities) and 12.4 Tg/yr (due to DMS) (Kiehl and Rodhe, 1995, Liaoa et al., 2003) (Table 1.1). Sulfate aerosols have a cooling effect as they are efficient scatterers of solar radiation.

52 52 Table 1.1: Global emission estimates for the main types of atmospheric aerosols. Source: Seinfeld and Pandis (2006) Aerosol Size Distribution The size of aerosols plays an important part in their lifetime and chemical interactions in the atmosphere. The task of defining the size distribution of a group of aerosols is extremely difficult, due to the large number of particles and the different size ranges that may characterize them, which cover several orders of magnitude. Therefore, in order to describe an aerosol size distribution, mathematical methods are to be used. In the following a brief description is presented for the derivation of the aerosol number distribution, n N (D p ), (Seinfeld and Pandis, 2006). The number of aerosol particles, per cm -3 of atmospheric air, with diameters ranging from D p to (D p + dd p ) is given by: n ( D ) dd (1.62) N p p

53 53 while the total number of aerosol particles, per cm -3 of atmospheric air, is: N n ( D ) dd. (1.63) 0 t N p p Due to the large range of diameters, characterizing the aerosol population, the number distribution is usually provided as a function of the logarithm of diameter and not the diameter itself, in which case the aerosol number distribution can be written as: n N (ln D ) d ln D (1.64) p p and the total number of particles is: N t n N (ln D ) d ln D (cm -3 ). (1.65) p p The observed aerosol size distribution can be well reproduced using the lognormal distribution (Aitchison and Brown, 1957). The normal distribution, for a parameter u, ranging from - to +, it is written as: 2 ( N ( u u) n u) exp( ) (1.66) 1/ 2 (2 ) 2 2 u u 2 with ū being the mean value of the distribution and u is the variance. The total of the parameter u is subsequently given by: N n( u) du. (1.67)

54 54 Figure 1.8: Normal distribution graph. Source: Finlayson-Pitts and Pitts (1986). The characteristic shape of a normal distribution resembles that of a bell, with the maximum value being at ū (Figure 1.8). The square root of the variance, which is the standard deviation, σ u, characterizes the distribution s width. In a normal distribution, 68% of the area located below the curve, falls within the range of ū ± σ u. In order for a parameter to have a lognormal distribution, the logarithm of the parameter has to have a normal distribution, therefore in the case of aerosols, their size distribution has to satisfy the equation (1.66), which is written as: n 2 dn N (ln D p ln D ) t pg ) exp( ), (1.68) 1/ 2 dd (2 ) D ln 2ln ( N D p 2 p p g g where N t, respectively. D pg and σ g are the total number of aerosols, the mean diameter and the standard deviation, Radiative Effects Aerosols affect the shortwave and longwave solar radiation by scattering and absorbing processes, changing the energy budget of the Earth-atmosphere system. This is known as the aerosol direct radiative effect. The indirect effect is the change that they induce to cloud properties. Due to the presence of aerosols, the size of cloud particles is smaller, than in clear conditions, with the number of them increasing, in order for the water content to be the same. A larger number of cloud droplets gives, as a result, a higher reflecting

55 55 efficiency to the cloud. The cloud lifetime is also affected, as the smaller particles lower the precipitation efficiency of the cloud and increase its life span. However, under specific conditions, increases of aerosol number concentrations may lead to increased precipitation (Khain et al., 2005, Lynn et al., 2005, van den Heever et al., 2006). The aerosol semi-direct effect is the changes in cloud cover and liquid water path caused by the absorption of solar radiation from aerosols. These effects are depicted in Figure 1.9. Figure 1.9: Illustration of aerosol radiative effects. Source: IPCC (2007). The radiative effects of aerosols are difficult to quantify with high accuracy, due to the high spatiotemporal variations that they present. The parameters that define the aerosols radiative effects are the optical properties, such as the single scattering albedo, phase function and the specific extinction coefficient, as well as the variation of these parameters with wavelength. These optical properties exhibit different values between the various types of aerosols, depending on each type s strength at scattering or absorption of solar radiation. Complicating the situation, aerosols found in the atmosphere rarely represent a single type, rather than a mixture of them, depending on the geographical and atmospheric conditions. The total aerosol burden in the atmosphere, as it is provided from the results of the ECHAM/GRANTOUR model (IPCC, 2001) can be seen in Figure Aerosol optical depth (AOD)

56 presents a high spatial distribution and receives the highest values in areas where dust emissions are dominant, as well as biomass burning and industrial areas. 56 Figure 1.10: Total optical depth of aerosols, based on the results of the ECHAM/GRANTOUR model. Source: IPCC (2001). Figure 1.11 presents the radiative forcing of the atmospheric constituents, as it calculated for the time between the years 1750 and 2005 (IPCC). The forcings are given in units of W m -2, along with their uncertainties and the current level of scientific understanding. The forcing of climate that is known with the highest accuracy is that of greenhouse gases, which are responsible for warming the system. The main cooling factor is provided by the effects of aerosols, which however hold a low accuracy in their estimated values. The aerosol direct effect is better understood than the indirect effect, which however could provide the highest magnitude of cooling for the Earth-atmosphere system. Aerosol effects could counteract the warming from the greenhouse gases, but their influence is more spatially restricted, whereas greenhouse gases are more homogeneously distributed. More studies are needed, taking advantage of ground-based, satellite, as well as model-derived results, in order to better understand the radiative forcing that the various atmospheric constituents contribute.

57 57 Figure 1.11: Radiative forcing of climate between 1750 and Source: IPCC (2007) Clouds Clouds are the major factor that affects the solar radiation variability on a diurnal basis. Temporal variations of cloudiness define the available incident solar radiation over an area. Clouds reflect the incoming shortwave solar irradiance, leading to a cooling in the Earth-atmosphere system and at the same time absorb the longwave irradiance emitted from the Earth and contribute to the warming created from the greenhouse gases. The effect of clouds on the solar irradiance depends, amongst others, on the cloud type, the size of cloud droplets and the height of the cloud. A detailed description about cloud properties and

58 radiative effects is provided, amongst others, by Liou et al. (2002), Vardavas and Taylor (2007) and Salby (2012). A brief description of the main aspects is provided in the following Cloud Types The three main types of clouds are stratiform, cumuliform and cirriform clouds. Stratiform clouds have a layered structure and are formed from the lifting of a stable layer, on a large scale. The lifetime is short, of the order of a day and they can be further divided into stratus, stratocumulus and nimbostratus, depending on the vertical extension. Cumuliform clouds are formed due to isolated updrafts. They have higher optical depth than stratiform clouds, their lifetime ranges from several minutes to hours and can be divided into cumulus and cumulonimbus clouds. Finally, cirriform clouds can be formed from either a largescale or an isolated air lifting and are present at higher altitudes. They mainly contain ice particles, in contrast with lower altitude stratiform and cumuliform clouds which are composed mostly by water particles. Table 1.2 presents the microphysical properties, depending on the cloud type, where n, α, ρ 1, ˆ e, β e are the cloud number density, mean droplet radius, liquid water content, mean extinction cross section and mean extinction coefficient, respectively. PSC I and PSC II, refer to Polar Stratospheric Clouds of Type I and II, which are a special category of clouds, formed over high latitudes under very cold temperature conditions. Data in this table are provided from the works of Carrier et al. (1967), WMO (1988), Dowling et al. (1990), Liou (1990), Knollenberg et al. (1993) and Turco et al. (1993). Stratus and cumulus clouds have large number densities but very small droplet radii, compared to cirrus and their extinction coefficient is in turn a few orders of magnitude higher. The different microphysical properties of each type of clouds, indicate the different strength that they have in the radiative forcing of the climate.

59 59 *Ice water content Table 1.2: Microphysical properties of main cloud types. Source: Salby (2012) Radiative Effects Figure 1.12: Refractive indices of water and ice, as a function of wavelength. Source: Fu (2002). Figure 1.12 presents the real (A) and imaginary (B) refractive indices of water and ice as a function of wavelength (Fu, 2002), in order to study the scattering and absorption of solar radiation by water and ice cloud particles, respectively. The real part of the refractive index, m r is almost the same for both types of

60 60 particles, in the visible part of the spectrum, where is shows no variation with wavelength, an effect that accounts for the white colour of clouds. As the wavelength increases, m r, no longer has a constant value and after 10 μm, water and ice particles begin to show variations between them. The imaginary part of the refractive index, m i, has very low values, up until around the wavelength of 10 μm, after which the behaviour of water and ice begins to differ. Absorption is higher for water particles in large infrared wavelengths. As mentioned before, clouds affect the shortwave and longwave irradiance and the effects depend on the cloud type, height, zenith angle of the sun etc. A quantitative description of these effects is provided in the following (Shalby, 2012). Considering the surface albedo, A, the heating rate over an area is equal to the difference of the absorbed and emitted energy flux, Q ( 1 A) F F ( z ), (1.69) S LW where F S and F LW are the downward shortwave flux and upward longwave flux at the top of the atmosphere. The radiative forcing, due to clouds, C, is equal to the difference of this heating rate and the heating rate considered under cloud-free conditions, CS Q C CS Q Q, (1.70) which, taking into account equation (1.69), is written as: CS CS C A A) F ( F F ), (1.71) ( S LW LW where CS denotes the cloud-free conditions. The cloud radiative forcing is divided into a shortwave and a longwave component, C SW and C LW, respectively: C C SW LW CS ( A A) F CS LW F LW F S F CS SW F SW (1.72) Considering η to be the cloud cover, the radiative fluxes are now written as:

61 61 F F SW LW (1 ) F (1 ) F CS SW CS LW F F OC SW OC LW (1.73) where OC denotes the overcast conditions. The components of the cloud radiative forcing are now given by: C C SW LW ( F ( F CS SW CS LW F F OC SW OC LW ) F ) S ( A CS A OC ) (1.74) Based on data collected from the Earth Radiation Budget Experiment (ERBE) instruments, on satellites ERBS and NOAA 9, the global annual shortwave and longwave cloud radiative forcing have values of -48 and 31 W m -2, respectively (Harrison et al., 1990), with the corresponding net cloud radiative forcing being equal to -17 W m -2. As a conclusion, the net radiative effect of clouds is one of cooling for the Earth-atmosphere system, although it has to be considered having a regional significance.

62 62 References Aitchison, J., Brown, J.A.C., The Lognormal Distribution Function. Cambridge University Press. Andrews, D.G., An Introduction to Atmospheric Physics, Second Edition. Cambridge University Press. Carrier, L.W., Cato, G.A., von Essen, K.J., The backscattering and extinction of visible and infrared radiation by selected major cloud models. Applied Optics 6, 7, Chung, S.H., Seinfeld, J.H., Global distribution and climate forcing of carbonaceous aerosols. Journal of Geophysical Research 107, D19, Dowling, D., Radke, R.R., Lawrence, F., A summary of the physical properties of cirrus clouds. Journal of Applied Meteorology 29, Finlayson-Pitts, B.J., Pitts, Jr., J.N., Atmospheric Chemistry: Fundamentals and Experimental Techniques. John Wiley and Sons, New York. Fu, Q., Radiative Transfer Cloud-Radiative Processes. Encyclopedia of Atmospheric Sciences (Edited by J. Holton, J. Pyle and J. Curry), Academic Press, Gong, S.L., Barrie, L.A., Lazare, M., Canadian Aerosol Module (CAM): A size-segregated simulation of atmospheric aerosol properties for climate and air quality models 2. Global sea-salt aerosol and its budgets. Journal of Geophysical Research 107, D24, Hansen, J.E., Travis, L.D., Light scattering in planetary atmospheres. Space Science Reviews 16, Harrison, E.F., Minnis, P., Barkstrom, B.R., Ramanathan, V., Cess, R.D., Gibson, G.G., Seasonal variation of cloud radiative forcing derived from the Earth Radiation Budget Experiment. Journal of Geophysical Research 95,

63 63 Intergovernmental Panel on Climate Change (IPCC): Climate Change 2001: The Scientific Basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change. [Houghton, J.T., Ding, Y., Griggs, D.J., Noguer, M., van der Linden, P.J., Dai, X., Maskell, K., Johnson, C.A., (eds)]. Cambridge University Press. Intergovernmental Panel on Climate Change (IPCC): Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. [Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M., Miller, H.L., (eds)]. Cambridge University Press. Khain, A., Rosenfeld, D., Pokrovsky, A., Aerosol impact on the dynamics and microphysics of deep convective clouds. Quarterly Journal of the Royal Meteorological Society 131, Kiehl, J.T., Rodhe, H., Modeling geographical and seasonal forcing due to aerosols. Aerosol Forcing of Climate, R.J. Charlson and J. Heintzenberg, eds., Wiley, New York, Knollenberg, R.G., Kelly, K., Wilson, J.C., Measurements of measurements of number densities of ice crystals in the tops of tropical cumulonimbus. Journal of Geophysical Research 98, Liao, H., Adams, P.J., Seinfeld, J.H., Mickley, L.J., Jacob, D.J., Interactions between tropospheric chemistry and aerosols in a unified GCM simulation. Journal of Geophysical Research 108, D1, Liao, H., Seinfeld, J.H., Adams, P.J., Mickley, L.J., Global radiative forcing of coupled tropospheric ozone and aerosols in a unified general circulation model. Journal of Geophysical Research 109, D16207, doi: /2003jd Liou, K., Radiation and Cloud Processes in the Atmosphere. Oxford University Press, New York. Liou, K.N., An Introduction to Atmospheric Radiation, Second Edition. International Geophysics Series 84, Academic Press. Liousse, C., Penner, J.E., Chuang, C., Walton, J.J., Eddleman, H., Cachier, H., A global threedimensional model study of carbonaceous aerosols. Journal of Geophysical Research 101,

64 64 Lynn, B.H., Khain, A.P., Dudhia, J., Rosenfeld, D., Pokrovsky, A., Seifert, A., Spectral (bin) microphysics coupled with a mesoscale model (MM5). part I: Model description and first results. Monthly Weather Review 133, Salby, M.L., Physics of the Atmosphere and Climate. Cambridge University Press. Seinfeld, J.H., Pandis, S.N., Atmospheric Chemistry and Physics: From Air Pollution to Climate Change, Second Edition. John Wiley and Sons, Inc. Thomas, G.E., Stamnes, K., Radiative Transfer in the Atmosphere and Ocean. Cambridge Atmospheric and Space Science Series, Cambridge University Press. Turco, R., Drdla, K., Tabazadeh, A., Hamill, P., Heterogeneous chemistry of polar stratospheric clouds and volcanic aerosols. The Role of the Stratosphere in Global Change, NATO ASI Series I, 8, M. Chanin, Ed., Springer Verlag, Heidelberg. van den Heever, S.C., Carrio, G.G., Cotton, W.R., DeMott, P.J., Prenni, A.J., Impacts of nucleating aerosols on Florida Storms. part I: Mesoscale Simulations. Journal of Atmospheric Science, doi: /jas Vardavas, I.M., Taylor, F.W., Radiation and Climate. International Series of Monographs on Physics 138, Oxford University Press, Oxford. Whitby, K.T., Cantrell, B., Fine particles. Proc. Int. Confe. Environmental Sensing and Assessment, Las Vegas, NV, Institute of Electrical and Electronic Engineers. WMO, Report of the International Ozone Trends Panel Report No. 18, World Meteorological Organization, Global Ozone Research and Monitoring Project, NASA, Washington, DC. Zender, C.S., Bian,H., Newman, D., Mineral dust entrainment and deposition (DEAD) model: Description and 1990s dust climatology. Journal of Geophysical Research 107, D24, 4416.

65 65 2. Data sources from remote sensing instruments and models In this chapter, the main atmospheric data sources, that were used during the work for this thesis, are presented. These include the ground-based network of AERONET, data retrieved from the satellite instruments SEVIRI and MODIS and data provided from the AeroCom project, which comprises of results from multiple aerosol modules. A description of the libradtran radiadive tranfer library, used throughout the presented thesis, is provided at the end of the chapter The AERONET network The Aerosol Robotic Network (AERONET) ( (Holben et al., 1998, 2001) is a ground-based network with stations around the globe, established by NASA and PHOTONS, dedicated to the study of atmospheric aerosols. In Figure 2.1, the distribution of the current operating AERONET stations is presented. The number of stations provides a good coverage over land, with the exception of parts of Russia, Sahara and Western Africa. The number of stations covering ocean territories is small, due to the difficulties and cost of maintaining a remotely located site.

66 66 Figure 2.1: The AERONET stations distribution. Source: The network provides data of AOD, inversion products and precipitable water. The data available from AERONET are provided under 3 different levels: Level 1.0, which are unscreened for cloud contamination, Level 1.5, which are cloud-screened and Level 2.0, which are cloud-screened and qualityassured. More specifically, the Level 2.0 data are pre- and post-field calibrated, automatically cloudscreened and afterwards inspected manually. Calibration is performed on the measuring instruments before they are placed in the field and after a measuring time period. The cloud-screening algorithm used is based on criteria applied on the AOD variability over a short-time (1 min) as well as an hourly and diurnal time period. The Level 2.0 data are finally checked manually in order to identify possible abnormal points.

67 67 Figure 2.2: The Cimel spectral radiometer, used by the AERONET network. Source: The AERONET measuring system comprises of Cimel spectral radiometers (Figure 2.2), performing direct sun and sky radiance measurements in order to retrieve aerosol optical properties. The direct sun measurements are performed at 8 wavelengths, 340, 380, 440, 500, 670, 870, 940 and 1020 nm, with the channel at 940 nm being used to derive the water column. The AOD at each wavelength is derived from the extinction of the direct beam, as given by the Beer Lambert Bouguer law, after taking into account the Rayleigh scattering of the atmospheric gases and the ozone absorption, V ( ) V ( )* 2 d *exp ( ) * T m, (2.1) 0 where V(λ), V 0 (λ) are the measured voltages at the surface and at the top of the atmosphere respectively, τ(λ) Τ is the total optical depth, m is the optical air mass and d is the ratio between the mean and the real value of the Earth Sun distance. The optical depth corresponding to aerosols (τ(λ) aerosol ) is derived from the total optical depth after taking into account the Rayleigh scattering and the absorption of the most important atmospheric gases, ( ) aerosol ( ) T ( ) Rayleigh ( ) water ( ) O CH, (2.2) ( ) ( ) ( ) 3 CO 2 NO 2 4 where τ(λ) Rayleigh, τ(λ) water, τ(λ) O3, τ(λ) CO2, τ(λ) NO2 and τ(λ) CH4 are the optical depths due to the Rayleigh scattering by the atmospheric gases and the absorption due to water vapor, ozone, carbon dioxide, nitrogen dioxide and methane. Direct sun measurements are performed every 15 minutes at each wavelength, during

68 68 the day and with a higher frequency during the time of sunrise and sunset. Each measurement is in fact a group of three measurements (triplet) taken within a minute with a 30 time difference between them. The variability of the triplet values is an indicator of the presence of clouds, since the time variability of clouds is higher than the one describing aerosols, therefore under clear-sky conditions the differences between the three measurements, in the 1 minute time frame under which they are acquired, should be very small. In order to derive the water vapor in the total atmospheric column, three channels are used at 675, 870 and 940 nm. After estimating the Rayleigh and aerosol extinction, the total transmission is estimated at the first two wavelengths, 675 and 870 nm. Subsequently, the total transmission at 940 nm is calculated with the method of extrapolation. The retrieval of the water vapor transmission is achieved by deducting this extrapolated water vapor transmission with the one measured at 940 nm and afterwards the precipitable water is calculated based on the following equations ln lnt 940extrapolated 2 T lnv * d lnv m * ( T W ln T940measured ln u ln T W W lnt a m W W o940 a * ( m 1/ b W * u) b AOT 940 Rayleigh ), (2.3) where u is the precipitable water in cm, T W is the water vapor transmission, a and b are constants depending on the filter and m W is the optical air mass of water vapor (Schmid, 2001, Smirnov, 2004). The sky radiance measurements are performed at 4 spectral bands, 440, 670, 870, 1020 nm and processed according to the Dubovik and Nakajima inversions in order to derive the size distribution and phase function of aerosols (Dubovik and King, 2000, Dubovik et al., 2000, 2002a, 2002b, 2006, Nakajima et al., 1996 Sinyuk et al., 2007). The measurements are performed along the solar principal plane and the solar almucantar up to 9 and 6 times a day respectively. The version 1.0 inversion code, which provided aerosol properties using a spherical and a spheroidal aerosol model, has been replaced with the version 2.0 code, which provides one set of aerosol data along with the percentage of spherical particles. Another important

69 change is the upgrade of surface albedo information, which is now taken from satellite and model estimations and the new criteria applied which ensure the provision of quality-assured Level 2.0 data. 69 Inversion products provided by the AERONET database include size distribution, complex refractive index, phase function, single scattering albedo, asymmetry parameter, absorption optical depth as well as spectral and broadband fluxes and estimations of the aerosol radiative forcing and radiative forcing efficiency. In order to model the sky radiance, the radiative transfer equation needs to be solved. For a planeparallel atmosphere, the angular distribution of diffuse downward radiation is given by I( ; ) F m I( ; ) F exp( m0 ) exp( m1 ) *( 0P( ; ) G(...)), 0 m m m exp( m )*( P( ; ) G(...)), 0, (2.4) where I (Θ;λ) is the sky radiance at wavelength λ and scattering angle Θ, F 0 is the radiative flux at the top of the atmosphere, m is the air mass (m 0 1/cosθ 0, m 1 1/cosθ), θ and θ 0 are the observation and solar zenith angles respectively, τ is the optical depth due to extinction, ω 0 is the single scattering albedo and P (Θ;λ) is the phase function. The multiple scattering effects are described by the expression G ( ) = G (ω 0 (λ); τ ext (λ); P (Θ; λ); A (λ); θ 0 ; θ; φ), where Α (λ) is the surface reflectance at wavelength λ and φ is the observation azimuth angle. The atmosphere is considered to be homogeneous, with no polarization effects being taken into account and the Lambertian approximation is considered for the surface reflectance. The discrete ordinate codes by Nakajima and Tanaka (1988) and Stamnes et al. (1988) are used to take into consideration the multiple scattering. The atmosphere is divided into homogeneous layers, each characterized by different values for the optical thickness, the phase function and the single scattering albedo. In order to define these three parameters, the gaseous absorption, molecular scattering and the scattering and absorption by atmospheric aerosols need to be considered. These parameters are then given as

70 70 total ext P total 0 total ( ) ( ) aer scat aer scat ( ; ) ( ) ( ) aer scat total scat total ext total abs mol scat ( ) ( ) P ( ) ( ) ( ) aer mol scat total scat total ext ( ; ) ( ) ( ) ( ) mol scat total scat gas abs ( ) P ( ) ( ) mol ( ; ), (2.5) aer aer aer where ext ( ), 0 ( ), P ( ; ) are the aerosol optical thickness, single scattering albedo and phase function of the layer, respectively. The gaseous absorption can be avoided by choosing the right measuring instrument and in the case of AERONET, the wavelengths at which the sky radiance measurements are performed, were chosen as such for this very reason. The molecular scattering can be derived from the surface pressure and a minor absorption from ozone can be considered using climatological data. Therefore the atmospheric radiance I (Θ;λ) is considered to vary, depending mainly on the aerosol optical properties and can now be written as aer aer aer I ( ; ) I( ( ); 0 ( ); P ( ; )). (2.6) ext Multiple radiance measurements in different wavelengths and scattering angles are needed for the retrieval of the three parameters characterizing atmospheric aerosols. Since the radiance measurements that are acquired from ground instruments are affected by the entire atmospheric column, these aerosol optical properties are calculated in the total atmospheric column, even though there are vertical variations. The aerosol optical parameters can be modeled using microstructure parameters (King et al., 1978, Nakajima et al., 1983, 1996). The optical thickness, due to aerosol extinction, scattering, absorption and the aerosol phase function, can be described as scat ( ) P( ; )...( ) r r max min K... r r max min ~ K scat ( ; m; r) n( r) dr ~ ( ; ; m; r) n( r) dr, (2.7) where r represents the particle radius, n(r) = dn(r)/dr is the particle number size distribution, K scat ( ) and K τ ( ) are a scattering and an extinction cross section respectively. The particles are considered to be

71 spherical, therefore the scattering and extinction cross section are approximated by Mie functions, corresponding to spherical and homogeneous particles with a complex refractive index given as 71 ~ m( ) n( ) ik( ). (2.8) As a result, instead of the optical thickness, single scattering albedo and phase function of aerosols, one can consider the size distribution and refractive index and the atmospheric radiance is now given by ~ I ( ; ) I ( dn ( r ) / dr ; m ( )). (2.9) The inversion, in order to retrieve the desired aerosol parameters, is performed by finding the best fit of all the data used by a theoretical model. Consideration is given to the different errors characterizing each set of data The AeroCom Project AeroCom is an international project ( initiated in 2003, aiming to the better understanding of the modeling of aerosols and their role in Earth-atmosphere interactions. The purpose of the project is the gathering and intercomparison of results from more than 14 aerosol models, along with ground measurements and satellite retrievals, in order to improve the design of aerosol modules in the climate models. Remote sensing information includes data from instruments on board satellites (MODIS, MISR, AVHHR, TOMS, POLDER, SeaWiFFS) and ground stations (AERONET). Uncertainties in model estimations regarding aerosol properties and radiative effects remain high and AeroCom aims in identifying the uncertainty sources among the different global models that are used and adopt more accurate insert data with regards to aerosol emission sources and sink mechanisms, interaction assumptions and optical properties. The first assessment for the AeroCom project was performed by comparing the results of 20 aerosol modules, regarding the mass global distribution and aerosol optical thickness at 550 nm (Kinne et al., 2006).

72 72 The results concern the so-called Experiment A conditions, where the aerosol models are run with their standard configurations, taking into concern either a climatological dataset from a time period of 3 to 10 years or meteorological datasets for the years 1996, 1997, 2000 or The models that were used are shown in Table 2.1, along with the time period considered and results that are provided. Table 2.1: Global aerosol models, taking part in the AeroCom project. Source: Kinne et al. (2006). The global aerosol optical thickness results, provided by the models, has values between 0.11 and 0.14, when the average value retrieved from AERONET and satellites measurements is and 0.15, respectively. The differences become more important when the aerosol distribution and compositional mixture is considered, especially in the cases of dust and carbonaceous aerosol. This, points to uncertainties between the models, concerning the absorption properties of aerosols. These differences induce larger uncertainties in the modeling of the radiative effects. Experiment B was performed, using the same emission sources for all models (Dentener et al., 2006), produced for the year 2000 as well as meteorological datasets for the same year. The results from 12 models were used for the intercomparison, showing small improvement in the agreement between the

73 73 different models products (Textor et al., 2007). This indicates that the emission rates and properties have a secondary role to the uncertainties produced in the estimation of aerosol optical properties and load. The main source of uncertainty lies with the transport and removal processes, considered by each model, as well as the chemical processes that lead to the creation of secondary particles and the parameterizations used to represent aerosol microphysics. The life cycles of aerosols, as they are simulated by 16 models participating in the AeroCom project, is investigated by Textor et al. (2006). The aerosol types that are investigated are mineral dust, sea salt, sulfate, black carbon and particulate organic matter. The life cycle of an aerosol type, as simulated by a model, depends on the model assumptions regarding sources and sinks, transport, interactions, particle size, water uptake, vertical and horizontal dispersal. The emissions of atmospheric particles are higher, in mass, for sea salt and mineral dust, while these two aerosol types are also the ones responsible for the highest burdens in the atmosphere, as the models results show. Black carbon and particulate organic matter seem to have the highest residence times and along with sulphates are removed primarily due to wet deposition. Dry deposition is mainly responsible for the removal of larger particles, like mineral dust and sea salt. The highest heights in the atmosphere are reached by black carbon, particulate organic matter and mainly sulphate aerosols. The diversities of aerosol life cycles, found between the models, mainly concern the size of the particles, emissions, residence times and deposition. The composition of the particles and the hygroscopic growth mechanisms, due to water uptake, also create differences between the models results. Finally, important diversities seem to concern the aerosol transport, as it is simulated by each model and the dispersal of aerosols, on a vertical level. The vertical distribution of aerosols, as it is simulated by AeroCom global models, is compared with profiles measured by lidars, in Ferrare et al. (2006). The AeroCom models derived aerosol optical thickness and extinction profiles are compared with measurements from the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Climate Research Facility (CRF) Raman lidar (CARL), for the period The agreement is quite good above 2 km, but in lower altitudes the model profiles are lower by around 30 to 50% from the lidar measurements. The models provide lower vertical variability in

74 74 the produced aerosol profiles. Another simulation, produced by the Global Ozone Chemistry Aerosol Radiation and Transport (GOCART) model, which was performed during the Transport and Chemical Evolution over the Pacific (TRACE-P) was compared with measurements from the NASA Langley Research Center airborne Differential Absorption Lidar (DIAL), onboard the NASA DC-8 aircraft. The profiles provided by GOCART showed a 10 to 40% underestimation of the DIAL data, with the greatest differences found around 1 km altitudes. The aerosol global models, participating in the AeroCom project, have been compared with regards to their simulation of desert dust properties, the corresponding radiative effect and deposition mechanisms (Huneeus et al., 2011). The results of 15 models were compared to observations, with the focus being on the AOD and dust deposition data, as well as the Angström exponent, coarse mode AOD and surface concentrations of mineral dust. Important differences appear to be between the models simulation of the dust cycle. AOD and Angström exponent, for mineral dust, when vertically integrated, are simulated within a factor of two from the global models, while total deposition and concentrations on the surface are simulated within a factor of 10. An overestimation is found the simulation of wet deposition in areas where the dry deposition is dominant. The seasonal variability and magnitude of AOD and Angström exponent is better simulated at stations located in Africa than in the Middle East. The dust transport across the Atlantic is well simulated during the summer than in winter or spring. Further measurements and comparisons are needed in order to improve the dust distribution, transport and deposition, as simulated by global models. An evaluation of AeroCom models was performed for another aerosol type, by Koch et al. (2009). Model simulations were compared with ground and aircraft measurements, with regards to black carbon particle burdens. Aerosol absorption optical depth from AERONET and Ozone Monitoring Instrument (OMI) were used for the evaluation, along with black carbon column data from AERONET. The majority of the models presented a high bias in surface concentration, for all areas except Asia. When considering the aerosol absorption optical depth however, the bias was lower. In South American and African regions, where biomass burning takes place, the ratio of modeled to retrieved data, for aerosol absorption optical depth, was found to be lower than 0.7 and 0.6, respectively. In Asian regions, the ratio was around 0.7.

75 Models show an underestimation of the absorption due to black carbon aerosols and further improvements in refractive indices, aerosol particle sizes and black carbon optical properties should be performed. 75 The aerosol direct radiative forcing has been estimated from the AeroCom project, based on simulations from 9 different models (Schulz et al., 2006). Prescribed aerosol emissions, one representing present day conditions (year 2000) and the other pre-industrial conditions (year 1750), were used, in order to derive the anthropogenic aerosol direct effect. The emission datasets, regarding primary aerosol and precursor gases, are described in detail in Dentener et al. (2006). Emission loads are provided, along with injection altitudes and particle sizes, with a spatial resolution of 1 x 1. Depending on the aerosol type, these data are provided on a daily (mineral dust, sea salt), monthly (particulate organic matter) or an annual basis (volcanic sulphate). The models did not take into consideration anthropogenic nitrate and dust, when simulating the direct aerosol effect. The radiative forcing at the top of the atmosphere, on a global annual basis and under all-sky conditions, was found to be Wm -2, with a range from to Wm -2 and a standard deviation of ±0.16 Wm -2. Larger values were found for the atmospheric and surface forcing (+0.82 Wm -2 and Wm -2, respectively). Local values can be significantly higher, especially when industrial regions are concerned. Uncertainties remain between the various models, even when the emission sources used as input are common, for all of them and in order to reach more accurate conclusions, regarding the aerosol direct effect, the mean results from multiple model outcomes should be considered. The aerosol direct radiative effect has been revisited, by Myhre et al. (2013), investigating the results of 16 global models, more than in Schulz et al. (2006). The models have undergone developments and upgrades during the years between the two studies. The total aerosol direct radiative forcing is found now to be more strongly negative than before. This is due to the fact that most models have taken into consideration the effects of nitrate and secondary organic particulates. The highest, in absolute values, radiative forcing appears to be the one caused by sulphate and black carbon particles. The results produced for sulphate don t present important differences from the previous ones but black carbon, produced from fossil and bio-fuel emissions, seems to have twice the strength than in Schulz et al. (2006). The mean value for the direct radiative forcing, calculated from the simulations of the 16 models, is found to be equal to Wm -2.

76 76 When performing certain modifications to the models that didn t take into account nitrates and secondary organic aerosols, the value is changed to Wm -2. The results in this study concern the period of 1850 to 2000, so if they are scaled to the period studied in Schulz et al. (2006), , the radiative forcing is then found to be equal to Wm -2, much stronger than estimated before. The aerosol indirect effects, as they are simulated by 10 models in the AeroCom project, are investigated by Quaas et al. (2009). The model outcomes are compared with observations from 3 satellite instruments (MODIS and CERES, onboard Terra and Aqua, and ATSR-2, on board ERS-2). The comparison is focused on stratiform liquid water clouds and the relationships between aerosols, clouds and radiation are studied. A positive correlation is found, between the cloud droplet number concentration and the AOD, in both satellite and models results. However, an overestimation is found in the models outcomes, over land. A model overestimation is also found in the relationship between cloud liquid water path and AOD, indicating the need for an improvement in the models cloud parameterization. The relationships between cloud fraction and AOD are found to be positive by the majority of the models, agreeing with the satellite retrievals, but are weaker than the latter Meteosat Second Generation (MSG) satellites The European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) ( was established in 1986 with the purpose of collecting and providing satellite data concerning weather and climate parameters, on a long-term basis. These satellite data are made available to the National Meteorological Services of the countries that are members of EUMETSAT. EUMETSAT is in charge of a fleet of satellites and the corresponding ground-based monitoring facilities, in order to continuously provide reliable data, satellite images and products. Figure 2.3 presents the EUMETSAT satellites, past, present and future, with their estimated operating timeline. EUMETSAT satellites are operational in an either geostationary (GEO) or Low Earth Orbit (LEO). The geostationary orbit is a geosynchronous orbit, which means that it is equal to the Earth s rotation

77 77 frequency, circular and at a fixed latitude of 0 degrees, on the equator plane. Meteosat Second Generation (MSG) satellites have come to replace the very successful Meteosat predecessors, who have been operating since 1977, bringing a new era to the meteorological studies and remote sensing of the Earth-atmosphere system. ( The MSG program includes four satellites, the first of which (MSG-1) was launched on 28/8/2002, followed by MSG-2 and MSG-3, on 22/12/2005 and 5/7/2012 respectively and MSG-4 in the near future. Once operational, the satellites have acquired the names Meteosat-8, Meteosat-9 and Meteosat-10, respectively.

78 78 Figure 2.3: EUMETSAT satellites with estimated lifetimes of operation. Sources: MSG mission is planned as a two-satellite project, where one is operational and the other is in orbit in a stand-by mode (Figure 2.4).

79 79 Figure 2.4: Presentation of the MSG operation mode. Sources: Figure 2.5: The Meteosat Second Generation (MSG) satellite. Sources: MSG is a spin stabilized satellite (Schmetzet al., 2002), spinning around its longitudinal axis in a counter-clockwise mode with a rate of 100 resolutions per minute (Figure 2.5). Its longitudinal axis is aligned with the rotational axis of the Earth. It is at a geostationary orbit, at a 0 degrees longitude, but can be moved 50 in the east or west direction if necessary. The imaging area covers Europe, Africa, the Atlantic

80 80 ocean, the Middle East and Eastern South America. MSG satellites are launched into space, carried by Ariane 5 rockets and put into an elliptical orbit, which is between a geostationary and a Low Earth orbit. The satellite consequently uses its own engine to gradually reduce the ellipticity of the orbit and finally achieve the geostationary orbit. The satellite has a height of 2.4 m, a diameter of 3.2 m and carries 2 scientific instruments. The Spinning Enhanced Visible and Infrared Imager (SEVIRI) and the Geostationary Earth Radiation Budget (GERB) are the main satellite detectors. The satellite s payload also includes a Search and Rescue system (S&R), monitoring distress signals and the Mission Communication Payload (MCP), which is the satellite transmitter, for providing the instrument s received data to the ground control stations. The GERB instrument is a visible-infrared radiometer dedicated to study Earth s budget by performing measurements of the incoming, to the Earth-atmosphere system, shortwave solar radiation and the outgoing longwave radiation. It is the first instrument onboard a geostationary satellite to perform such a study. The measurements include the solar band 0.32 to 4 μm and the total wavelength range of 0.32 to 30 μm. The longwave band is then calculated as the difference of the two (Dewitte et al., 2000). SEVIRI (Figure 2.6) performs observations at 12 different wavelength channels with a temporal resolution of 15 minutes. Out of the 12 channels, 11 have an image sampling distance of 3 km at nadir and one, the high-resolution visible channel (HRV) has an image sampling distance of 1 km at nadir. This is a big advancement, when compared to the first Meteosat satellites, which made observations at only 3 channels with a temporal revolution of 30 minutes, while the image sampling distance was 5 km and 2.5 km for the visible channel, respectively. The 11 channels make observations of the Earth s complete disk, while the HRV channel images half the disk in the East-West direction and the complete disk in the North-South direction.

81 81 Figure 2.6: The SEVIRI instrument onboard MSG. Sources: SEVIRI s telescope optical system includes three mirrors, with a main mirror of 51 cm in diameter. The temperature of the IR sensors is controlled and is kept at either 85 or 95 K. It takes 12 minutes for the imaging of the entire disk to be completed, after which the thermal IR channels are calibrated. Calibration is achieved with a blackbody that is positioned into the instrument s optical path. After 15 minutes, the procedure is ready to be repeated. The high temporal resolution of the imager, gives the opportunity to track rapid changes in cloud cover and optical properties, which as a result brings better understanding of the climate system interactions and radiative budget studies. SEVIRI s visible channels are at 0.6 and 0.8 μm and are used for cloud detection, aerosol and surface studies. The channel at 1.6 μm is used for aerosol studies and the distinction of snow and cloud pixels, as well as ice and water clouds. Low clouds and fog are identified from observations at 3.9 μm. Water vapor and wind observations are achieved with the measurements provided by the channels at 6.2 and 7.3 μm. The

82 82 infrared channels at 8.7 and 9.7 μm provide information regarding cirrus clouds and ozone, respectively. Finally infrared channels at 10.8, 12 and 13.4 μm are used, amongst others, for sea-land studies, ash clouds detection and cirrus clouds, respectively. The HRV channel is a broadband channel, gathering information in the μm wavelength range, used mainly for the water vapor studies. Level-1 data from SEVIRI are processed at ground facilities to provide level-1.5 image data. This primary process includes correction for the differences in detector response, as there are three detectors for each channel, except the HRV which has one, corrections for non-linear effects and finally geometrical correction. These level-1.5 data are further processed to retrieve all meteorological products. A cloud analysis is performed (Lutz, 1999) to identify the cloudy and clear pixels and retrieve information regarding cloud cover, cloud-top temperature and pressure, height etc. For cloud free pixels, the clear-sky radiance (W m -2 sr -1 (cm -1 ) -1 ) is calculated. Tropospheric humidity is calculated at two layers, the first between 850 and 300 hpa and the second between 500 and 200 hpa, based on information from the channels at 7.3 and 6.3 μm, respectively. Tracking successive images of cloud and water vapor patterns and obtaining information on wind speed and direction has as a result the production of the atmospheric motion vectors, a parameter of vital importance to numerical weather prediction. Other products include, total ozone, radiative fluxes, surface temperature, vegetation parameters, surface albedo etc MODIS satellite instrument The Earth Observing System (EOS) ( established and operated by NASA is a group of polar-orbiting and low inclination satellites, dedicated to observe the Earth-atmosphere system. Part of the EOS mission are the satellites Terra (EOS AM-1) ( and Aqua (EOS PM-1) ( launched on December 18, 2000 and May 4, 2002, respectively. Terra and Aqua satellites are designed to carry a sun-synchronous, near-polar, circular orbit at an altitude of 705 km. Terra crosses the equator at a descending mode, at a.m., while Aqua follows at an ascending mode, at 1.30 p.m., covering almost all of the Earth s surface every 1-2 days.

83 83 Figure 2.7: Terra satellite. Source: Terra s payload includes five scientific instruments to observe Earth s atmosphere, land and ocean (CERES, MISR, MODIS, ASTER, MOPITT) (Figure 2.7). These sensors provide information regarding atmospheric constituents, cloud and aerosol properties, energy budget and can monitor volcanoes, wildfires and other natural phenomena posing a threat, such as floods, storms and droughts. Aqua carries the instruments AIRS, AMSU, CERES, MODIS and AMSR-E, providing detailed information regarding the water cycle, atmospheric water vapor, precipitation, sea and land ice etc. Aqua was the first member of the Afternoon Constellation (A-Train), to be launched. A-Train is a group of satellites, given the corresponding name as they cross the equator at approximately 1.30 in the afternoon. There are currently 4 satellites of the A-Train operating, GCOM-W1, Aqua, ClouSat, CALIPSO and Aura (Figure 2.8).

84 84 Figure 2.8: The A-Train satellites. Source: The MODerate resolution Imaging Spectroradiometer (MODIS) ( onboard the two satellites, has been providing data since February 2000 and July 2002, for Terra and Aqua respectively (Figure 2.9). MODIS sensor acquires data in 36 spectral bands, extending from to μm. The spatial resolution at nadir is 250 km (for bands 1-2), 500 km (for bands 3-7) and 1 km (for bands 8-36). MODIS scan mirror scans ±55 degrees, providing a km swath path. The instrument s optical system redirects the received signal to 4 refractive objective assemblies, representing the four spectral regions (VIS, NIR, SWIR/MWIR, LWIR). MODIS provides information regarding the cloud cover and properties, aerosol parameters, water vapor in the atmospheric column, photosynthetic activity of plants, as well as tracks changes over the Earth s surface.

85 85 Figure 2.9: The MODIS instrument. Source: In order to retrieve aerosol information, two algorithms have been developed, one for retrievals over land (Kaufman et al., 1997a) and one for retrievals over ocean (Tanré et al., 1997). These algorithms have since been updated, leading to new versions, whereas the data available are grouped into collections depending on the process from which they have been produced. The newest collection of aerosol products, Collection 5 (C005) (Levy et al., 2007a, 2007b) has come to replace the previously used Collection 4 (C004) (Remer et al., 2005), making use of new surface reflectance data, aerosol model optical properties and lookup-tables, for retrievals performed over land. The MODIS reflectance data in three channels (0.47, 0.66 and 2.12 μm) are used to derive the spectral AOD and fine aerosol weighting (η) (Levy et al., 2007b). Reflectance data from other satellite channels are used for the cloud masking and pixel selection. The products mentioned are provided at the wavelength of 0.55 μm and at 10 km resolution. In order to derive the aerosol load over land, the methodology from Kaufman et al., 1997a is used. The reflectance at the top of the atmosphere (ρ * ) at a wavelength λ is given by: s * F ( 0 ) T ( ) ( 0,, ) ( 0,, ) ( 0,, ), (2.10) s 1 s (,, ) 0

86 86 where, T, s are the path reflectance, the angular surface reflectance, the normalized downward s, F d, flux when the surface reflectance is considered equal to zero (total downward transmission), the upward total transmission into the point of view of the satellite and the atmospheric backscattering ratio, respectively. These parameters are functions of the angles θ, θ 0 and φ, which are the zenith angle of the satellite, the solar zenith angle and the azimuth angle of the scattered radiation. This equation assumes a Lambertian surface. The surface reflectance is given the value corresponding to the solar-satellite viewing geometry, in order to account for the bidirectional reflectance properties. The path reflectance is a function of the aerosol optical thickness, single scattering albedo and phase function. The retrievals of AOD and fine aerosol weighting are achieved with the use of an aerosol look-up-table, that is examined in order to find the aerosol values that best produce the satellite reflectance signal. Both fine and coarse aerosol models are used for the inversion of the spectral reflectance, which is written again as a combination of the two:, (2.11) * * f *c ( 1) where are functions of the surface reflectance and path reflectance of the fine and coarse aerosol * f *c, model, respectively: * f * c af ac F F f d c d T f c s /(1 s f s ) s c s T /(1 s ), (2.12) The symbols are the same as in equation 1, with c and f indexes, describing the fine and coarse aerosol, respectively. The process performed, based on the algorithm over land, is illustrated in Figure 2.10 (Levy et al., 2007b). The information used for the retrievals include spectral reflectance and geolocation data. The spectral reflectances that are used, are those at 0.66, 0.86 μm (spatial resolution of 250 m), 0.45, 0.55, 1.24, 1.64, 2.12 μm (spatial resolution of 500 m) and 1.38 μm (spatial resolution of 1 km). The spatial resolution of the geolocation data is 1 km and these data include the zenith and azimuth angles of the satellite and sun, the scattering angle, the latitude, longitude, elevation and date of observation.

87 87 Figure 2.10: MODIS retrieval algorithm over land. Source: Levy et al. (2007b). A box of 400 pixels is analysed to find the pixels, where the aerosol retrieval is deemed applicable. Pixels that are cloud covered or include snow, ice or water covered areas are not taken into consideration. Dark pixels, with reflectance at 2.12 μm between 0.01 and 0.25, are chosen, out of which the brightest 50%

88 88 and darkest 20% are rejected. If the number of the remaining pixels is higher than 12, then the mean reflectance at channels 0.47, 0.66, 1.24 and 2.12 μm is calculated. The data of the look-up-table are corrected for the effect of the surface elevation. Depending on the location and time of the observation, a fine aerosol model is chosen. An interpolation is performed on the parameters ρ α, s, T and F with regards to the angles θ 0, θ and φ, which produces six values for each parameter, one for each value of the aerosol loading that is considered. For values of η, between -0.1 and 1.1 (step 0.1), the algorithm searches for the values of AOD at 0.55 μm and surface reflectance at 2.12 μm, that correspond to the given satellite reflectance at 0.47 μm. The solution that is chosen is the one that minimises the error, ε, produced at 0.66 μm, as described in the following equations: * ABS( 0.47 * ABS( 0.66 * ABS( 2.12 m 0.47 m 0.66 m 2.12 ) / ) / ) / m 0.47 m 0.66 m (2.13) If the reflectance at 2.12 μm is higher than 0.25, then procedure B is followed, but the retrieval is expected to have a lower accuracy. The new algorithm provides AOD retrievals within the expected accuracy level, which is equal to ±0.05±0.15τ The Radiative Tranfer Model LibRadtran is a software library consisting of radiative transfer codes and routines, originally designed by the collaboration of the German Aerospace Center (DLR) and the Norwegian Institute for Air Research (NILU) ( The software library was designed around the central program, the uvspec radiative transfer model (Mayer and Kylling, 2005). Other programs are included in the software library, in order to perform Mie calculations (mie), extract cloud properties (cldprp), calculate parameters regarding the sun s position (zenith, noon, sza2time), add levels to a given atmospheric profile (addlevel), perform interpolation, convolution and integration calculations (spline, conv, integrate) and other applications.

89 About uvspec Uvspec s original purpose was to carry radiative calculations in the UV and visible wavelengths ranges but it has since then been updated and now is used for many operations, like simulating scientific instruments, performing radiative budget calculations, used in remote sensing applications and many more. Uvspec is easy to use, as it reads an input file created by the user, where all the necessary atmospheric profiles and parameters are described, and provides the desired radiation field in an output file. The user can choose from multiple radiative transfer solvers, depending on the nature of the requested output data. A procedure is described by Stamnes et al. (1991) in order to derive total column ozone amount and cloud optical depth, from global irradiance measurements, using the libradtran software. Ozone is derived by comparing the ratio of irradiance values, measured at two wavelengths, one of which presents strong ozone absorption, with ratios of irradiance that have been calculated for multiple ozone amounts. Under cloud-free conditions, the results are very good, but as the cloud optical depth increases, the method overestimates the retrieved ozone values (Mayer et al., 1998). The cloud optical depth is derived, using the libradtran programs, by comparing irradiance measurements, at a wavelength with limited ozone absorption, with model calculations performed for multiple cloud optical depth values. (2005). The architecture of the uvspec model is described in Figure 2.11, taken from Mayer and Kylling

90 90 Figure 2.11: Illustration of the uvspec algorithm. Source: Mayer and Kylling (2005). The atmospheric profiles, concentrations of gases, surface albedo, aerosol and cloud properties etc. are defined by the user in the input file and consequently the corresponding cross sections and parameterizations are used for the calculation of the profiles of optical properties of the atmospheric constituents. A radiative transfer solver is chosen to derive the transmittance and radiances, which after processing result in the output irradiance values, actinic fluxes or other radiation quantities requested by the user. The model s wavelength resolution is defined by three grids, the input grid, the internal grid and the output grid. The internal grid is not chosen by the user, but by libradtan itself. The output grid is defined by

91 91 the wavelength resolution provided in the solar spectrum file. The internal grid that is used by LibRadtran, in the UV and visible wavelength range, uses a step of 0.5 nm for wavelengths lower than 350 nm and a step of 1 nm for wavelengths higher than 350 nm. This is chosen so that computational time is saved and the wavelength bands, where ozone absorption and extinction by molecules, aerosols and clouds is important, are taken into account. This is outlined in Figure 2.12, taken from Mayer et al. (1997) Figure 2.12: Extraterrestrial spectrum (top left), atmospheric transmittance (top right) and resulting spectrum at the surface (bottom). Source: Mayer et al. (1997). The transmittance (upper right figure) is derived on a moderate resolution grid and consequently interpolated to the high resolution grid, given by the solar spectrum used in the input file (upper left figure), and multiplied with the extraterrestrial irradiance values, giving the output grid of the received radiation (bottom figure).

92 Spectral Calculations and Radiative Transfer Solvers The spectral calculations, performed by uvspec, can be either one of four choices: spectrally resolved calculations, line-by-line calculations, correlated-k methods or pseudo-spectral calculations. The spectrally resolved calculation is the most popular choice, when the wavelength region examined is in the UV or visible spectrum. The most important gas absorption bands in this area are the bands of Hartley, Huggins and Chappuis, for ozone and the resolution of 1 nm, upon which the calculations are performed, is sufficient to provided accurate results. In infrared wavelengths, molecules present absorption located in a very large number of narrow lines and in order to perform calculations in this case, the line-by-line method is used. A spectrally resolved absorption cross section profile is defined by the user. Line-by-line is the most accurate method to use in infrared bands but it consumes a great amount of time and computational effort. Instead it is more common to use band parameterization and most particularly the correlated-k approximation. Correlated-k approximations are included in uvspec, provided by Kato et al. (1999), Fu and Liou (1992) and Kratz (1995). Three versions of Kato et al. (1999) are included: Kato, which considers 575 spectral subbands, Kato2, with 148 spectral subbands and absorption coefficients based on HITRAN 2000 and Kato2.96, with absorption coefficients based on HITRAN96. The correlated-k method is the most efficient method to perform accurate and fast solar radiation calculations. The method presents uncertainties in wavelengths higher than 2.5 μm, but these have a small contribution to the total irradiance and since the results are integrated over the entire solar spectrum, the uncertainties are not important. In correlated-k calculations, the wavelength grid is defined by uvspec and results are provided over spectral bands and not as a function of wavelength. The pseudo-spectral calculations is another option for calculations in the entire solar spectrum. Uvspec uses the absorption parameterization from LOWTRAN/SBDART (Ricchiazzi et al., 1998), which takes into consideration low-resolution band models, designed for LOWTRAN 7 (Pierluissi and Peng, 1985). The atmospheric transmission is provided for the wavelength range of 0 to cm -1, with a resolution of 5 nm in the visible and around 200 nm in the thermal infrared range.

93 93 Uvspec includes several solvers for the equation of the radiative transfer, the most important of which are: disort, the algorithm for plane-paraller atmosphere, described in Stamnes et al. (1988), sdisort, the pseudospherical disort algorithm, where a correction for the Earth s sphericity is applied (Dahlback and Stamnes, 1991), polradtran, the plane-parallel solver which takes into account polarization (Evans and Stephens, 1991) and twostr, the two-stream radiative transfer solver (Kylling et al., 1995) Atmospheric input data A data file of an atmospheric vertical profile is chosen, which includes at least three parameters: pressure, temperature and altitude. The file may also contain the profiles for oxygen, water vapor, ozone, carbon dioxide and other gases. Data are defined at specific levels instead of layers. The air density is calculated from the profiles of pressure and temperature. The six atmospheric profiles, described in Anderson et al. (1986), are provided by libradtran. These profiles are the afglms, afglmw, afglss, afglsw, afgus and afglt which correspond to midlatitude summer conditions, midlatitude winter, subarctic summer, subarctic winter, the U.S. standard profile and tropical conditions, respectively. The solar file defines the extraterrestrial spectrum and contains two columns with data, corresponding to the wavelength and the extraterrestrial flux at that wavelength. There are four main extraterrestrial flux files included in libradtran: Atlas2, ranging from 200 to 420 nm with a 0.05 nm step (Woods et al., 1996), Altas3 ranging from 200 to 407 nm with a 0.05 nm step (Woods et al., 1996), the Kurucz flux file, ranging from 250 to nm with a 0.1 or a 1 nm step (Kurucz, 1992$) and the Gueymard flux file, ranging from 0.5 to 10 6 nm with a variable step (Gueymard, 2003). Solar flux files, derived from the combination of 2 or more different files, are also provided by the model, like the atlas_plus_modtran file, which combines the Atlas3 file for the region nm, the Atlas2 for the region nm and Modtran3.5 for the wavelength region of nm. Aerosol can be defined in the uvspec input file with many parameters. The simplest description is provided by choosing the aerosol default option, which invokes the aerosol model by Shettle (1989). This

94 94 model considers rural type aerosol in the boundary layer, background type aerosol for heights above 2 km, spring-summer atmospheric conditions and a visibility equal to 50 km. The model configurations can be changed by descriptions provided by the user for certain aerosol properties. More specifically, the used may define the aerosol type observed in the lower 2 km (aerosol_haze), in heights above 2 km (aerosol_vulcan), the seasonal profile used regarding aerosols (aerosol_season) and the visibility (aerosol_visibility). More parameters are available for the user to choose, which will overwrite the ones set by the default model of Shettle (1989). The AOD, single scattering albedo and asymmetry function can be defined either by the integrated in the column value or by using a profile. The Ångström alpha and beta coefficients can be defined, to take into account the wavelength dependence of AOD, as well as the aerosol phase function. Aerosol microphysical properties are also an option in uvspec, by defing the complex refractive index and size distribution. Water and ice clouds are included in uvspec and can be defined by describing the vertical profiles of liquid and ice water and effective droplet radius. These parameters can be chosen to correspond to a specific level or atmospheric layer. Water and ice clouds are treated mostly the same way in uvspec, with the main differences being due to the fact that ice particles are non-spherical particles. This results in a number of parameterizations to be made, in order for the conversion between microphysical and optical properties to take place. The parameterizations include the ones by Fu (1996) and Fu et al. (1998), which are appropriate for calculating fluxes but not radiances. The optical properties used, are delta-scaled properties. The parameterization by Fu (1996) is used for wavelengths lower than 4 μm, while for those higher than 4 μm, Fu et al. (1998) is accepted. The parameterization by Key et al. (2002) is used for radiance calculations and takes into consideration the wavelength region of 0.2 to 5 μm. The output of uvspec can provide direct and diffuse irradiances and actinic fluxes at the surface or at specified altitudes if desired. The radiances at specific angles can also be chosen to be derived by the model. Uvspec results have been extensively validated against ground or airborne measurements and other model calculations (e.g. Bais et al., 2003, Cahalan et al., 2005, Kylling et al., 1993, Mayer et al., 1997) and if the

95 uncertainties describing both model and measurements are taken into account, then uvspec provides very good results for irradiance calculations. 95

96 96 References Anderson, G.P., Clough, S.A., Kneizys, F.X., Chetwynd, J.H., Shettle, E.P., AFGL Atmospheric Constituent Profiles (0-120 km). AFGL-TR , AFGL (OPI), Hanscom AFB, MA Bais, A.F., Madronich, S., Crawford, J., Hall, S.R., Mayer, B., van Weele, M., Lenoble, J., Calvert, J.G., Cantrell, C.A., Shetter, R.E., Hofzumahaus, A., Koepke, P., Monks, P.S., Frost, G., McKenzie, R., Krotkov, N., Kylling, A., Swartz, W.H., Lloyd, S., Pfister, G., Martin, T.J., Roeth, E.-P., Griffioen, E., Ruggaber, A., Krol, M., Kraus, A., Edwards, G.D., Mueller, M., Lefer, B.L., Johnston, P., Schwander, H., Flittner, D., Gardiner, B.G., Barrick, J., Schmitt, R., International Photolysis Frequency Measurement and Model Intercomparison (IPMMI): Spectral actinic solar flux measurements and modeling. Journal of Geophysical Research 108, doi: /2002jd Cahalan, R., Oreopoulos, L., Marshak, A., Evans, K., Davis, A., Pincus, R., Yetzer, K., Mayer, B., Davies, R., Acherman, T.P., Barker, H.W., Clothiaux, E.E., Ellingson, R.G., Garay, M.J., Kassianov, E., Kinne, S., Macke, A., O Hirok, W., Partain, P.T., Prigarin, S.M., Rublev, A.N., Stephens, G.L., Szczap, F., Takara, E.E., Varnai, T., Wen, G., Zhuraleva, T.B., The International Intercomparison of 3-D Radiation Codes (I3RC): Bringing together the most advanced radiative transfer tools for cloudy atmospheres. Bulletin of the American Meteorological Society 86, Dahlback, A., Stamnes, K., A new spherical model for computing the radiation field available for photolysis and heating at twilight. Planetary and Space Science 39, Dentener, F., Kinne, S., Bond, T., Boucher, O., Cofala, J., Generoso, S., Ginoux, P., Gong, S., Hoelzemann, J.J., Ito, A., Marelli, L., Penner, J.E., Putaud, J.-P., Textor, C., Schulz, M., van der Welf, G.R., Wilson, J., Emissions of primary aerosol and precursor gases in the years 2000 and 1750 prescribed data-sets for AeroCom. Atmospheric Chemistry and Physics 6, Dewitte, S., Clerbaux, N., Gonzalez, L., Hermans, A., Ipe, A., Joukoff, A., Sadowski, G., Generation of GERB unfiltered radiances and fluxes. Proceedings 2000 EUMETSAT Meteorological Satellite Data Users Conference, Bologna, Italy, EUMETSAT EUM P 29,

97 Dubovik, O., King, M.D., A flexible inversion algorithm for retrieval of aerosol optical properties from Sun and sky radiance measurements. Journal of Geophysical Research 105, Dubovik, O., Smirnov, A., Holben, B.N., King, M.D., Kaufman, Y.J., Eck, T.F., Slutsker, I., Accuracy assessment of aerosol optical properties retrieval from AERONET sun and sky radiance measurements. Journal of Geophysical Research 105, Dubovik, O., Holben, B.N., Lapyonok, T., Sinyuk, A., Mishchenko, M.I., Yang, P., Slutsker, I., 2002a. Nonspherical aerosol retrieval method employing light scattering by spheroids. Geophysical Research Letters 10, /2001GL Dubovik, O., Holben, B.N., Eck, T.F., Smirnov, A., Kaufman, Y.J., King, M.D., Tanré, D., Slutsker, I., 2002b. Variability of absorption and optical properties of key aerosol types observed in worldwide locations. Journal of Atmospheric Science 59, Dubovik, O., Sinyuk, A., Lapyonok, T., Holben, B.N., Mishchenko, M., Yang, P., Eck, T.F., Volten, H., Munoz, O., Veihelmann, B., van der Zander, W.J., Leon, J.F., Sorokin, M., Slutsker, I., Application of spheroid models to account for aerosol particle nonsphericity in remote sensing of desert dust. Journal of Geophysical Research 111, D11208, doi: /2005jd Evans, K.F., Stephens, G.L., A new polarized atmospheric radiative transfer model. Journal of Quantitative Spectroscopy and Radiative Transfer 46, Ferrare, R.A., Browell, E.V., Hair, J.W., Ismail, S., Turner, D.D., Clayton, M., Butler, C.F., Brackett, V.G., Fenn, M.A., Notari, A., Kooi, S.A., Chin, M., Guibert, S., Schulz, M., Chuang, C., Krol, M., Bauer, S.E., Liu, X., Myhre, G., Seland, Ø., Fillmore, D., Ghan, S., Gong, S., Ginoux, P., Takemura, T., The Vertical Distribution of Aerosols: Lidar Measurements vs. Model Simulations. Proceedings 23 rd International Laser Radar Conference, July, 2006, edited by N.S. Chikao Nagasawa, Nara, Japan. Fu, Q., Liou, K.N., On the correlated k-distribution method for radiative transfer in nonhomogeneous atmospheres. Journal of Atmospheric Sciences 49,

98 Fu, Q., An accurate parameterization of the solar radiative properties of cirrus clouds in climate models. Journal of Climate 9, Fu, Q., Yang, P., Sun, W.B., An accurate parameterization of the infrared radiative properties of cirrus clouds in climate models. Journal of Climate 11, Gueymard, C., The sun s total and spectral irradiance for solar energy applications and solar radiation models. Solar Energy 76, Holben, B.N., Eck, T.F., Slutsker, I., Tanré, D., Buis, J.P., Setzer, A., Vermote, E., Reagan, J.A., Kaufman, Y.J., Nakajima, T., Lavenu, F., Jankowiak, I., Smirnov, A., AERONET A federated instrument network and data archive for aerosol characterization. Remote Sensing of Environment 66 (1), Holben, B.N., Tanré, D., Smirnov, A., Eck, T.F., Slutsker, I., Abuhassan, N., Newcomb, W.W., Schafer, J., Chatenet, B., Lavenue, F., Kaufman, Y.J., Vande Castle, J., Setzer, A., Markham, B., Clark, D., Frouin, R., Halthore, R., Karnieli, A., ONeill, N.T., Pietras, C., Pinker, R.T., Voss, K., Zibordi, G., An emerging ground-based aerosol climatology: Aerosol Optical Depth from AERONET. Journal of Geophysical Research 106, Huneeus, N., Schulz, M., Balkanski, M., Griesfeller, J., Prospero, J., Kinne, S., Bauer, S., Boucher, O., Chin, M., Dentener, F., Diehl, T., Easter, R., Fillmore, D., Ghan, S., Ginoux, P., Grini, A., Horowitz, L., Koch, D., Krol, M.C., Landing, W., Liu, X., Mahowald, N., Miller, R., Morcrette, J.-J., Myhre, G., Penner, J., Perlwitz, J., Stier, P., Takemura, T., Zender, C.S., Global dust model intercomparison in AeroCom phase I. Atmospheric Chemistry and Physics 11, Kato, S., Ackerman, T.P., Mather, J.H., Clothiaux, E.E., The k-distribution method and correlated-k approximation for a shortwave radiative transfer model. Journal of Quantitative Spectroscopy and Radiative Transfer 62,

99 99 Kaufman, Y.J., Tanré, D., Remer, L.A., Vermote, E.F., Chu, A., Holben, B.N., 1997a. Operational remote sensing of tropospheric aerosol over land from EOD moderate resolution imaging spectroradiometer. Journal of Geophysical Research 102, D14, Key, J., Yang, P., Baum, B., Nasiri, S., Parameterization of shortwave ice cloud optical properties for various particle habits. Journal of Geophysical Research 107, D13, /2001JD King, M.D., Byrne, D.M., Herman, B.M., Reagan, J.A., Aerosol size distributions obtained by inversion of spectral optical depth measurements. Journal of Atmospheric Science 21, Kinne, S., Schulz, M., Textor, C., Guibert, S., Balkanski, Y., Bauer, S.E., Berntsen, T., Berglen, T.F., Boucher, O., Chin, M., Collins, W., Dentener, F., Diehl, T., Easter, R., Feichter, J., Fillmore, D., Ghan, S., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Herzog, M., Horowitz, L., Isaksen, I., Iversen, T., Kirkevåg, A., Kloster, S., Koch, D., Kristjansson, J.E., Krol, M., Lauer, A., Lamarque, J.F., Lesins, G., Liu, X., Lohmann, U., Montanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S., Seland, Ø., Stier, P., Takemura, T., Tie, X., An AeroCom initial assessment optical properties in aerosol component modules of global models. Atmospheric Chemistry and Physics 6, Koch, D., Schulz, M., Kinne, S., McNaughton, C., Spackman, J.R., Bond, T.C., Balkanski, Y., Bauer, S., Berntsen, T., Boucher, O., Chin, M., Clarke, A., De Luca, N., Dentener, F., Diehl, T., Dubovik, O., Easter, R., Fahey, D.W., Feichter, J., Fillmore, D., Freitag, S., Ghan, S., Ginoux, P., Gong, S., Horowitz, L., Iversen, T., Kirkevag, A., Klimont, Z., Kondo, Y., Krol, M., Liu, X., Miller, R., Montanaro, V., Moteki, N., Myhre, G., Penner, J.E., Perlwitz, J., Pitari, G., Reddy, S., Sahu, L., Sakamoto, H., Schuster, G., Schwarz, J.P., Seland, Ø., Stier, P., Takegawa, N., Takemura, T., Textor, C., van Aardenne, J.A., Zhao, Y., Evaluation of black carbon estimations in global aerosol models. Atmospheric Chemistry and Physics 9, Kratz, D.P., The correlated k-distribution technique as applied to the AVHRR channels. Journal of Quantitative Spectroscopy and Radiative Transfer 53,

100 Kurucz, R., Synthetic infrared spectra. Proceedings of the 154 th Symposium of the International Astronomical Union (IAU); Tucson, Arizona, March 2-6, 1992, Kluwer, Acad., Norwell, MA. 100 Kylling, A., Stamnes, K., Meier, R.R., Anderson, D.E., The 200- to 300-nm radiation field in the stratosphere: Comparison of models with observations. Journal of Geophysical Research 98, Kylling, A., Stamnes, K., Tsay, S.-C., A reliable and efficient two-stream algorithm for spherical radiative transfer; Documentation of accuracy in realistic layered media. Journal of Atmospheric Chemistry 21, Levy, R.C., Remer, L.A., Dubovik, O., 2007a. Global aerosol optical properties and application to Moderate Resolution Imaging Spectroradiometer aerosol retrieval over land. Journal of Geophysical Research 112, D13210, doi: /2006jd Levy, R.C., Remer, L.A., Mattoo, S., Vermote, E.F., Kaufman, Y.J., 2007b. Second-generation operational algorithm: Retrieval of aerosol properties over land from inversion of Moderate Resolution Imaging Spectroradiometer reflectance. Journal of Geophysical Research 112, D13211, doi: /2006jd Lutz, H.J., Cloud processing for Meteosat Second Generation. EUMETSAT Technical Departement Technical Memorandum, 4, 26pp. Mayer, B., Seckmeyer, G., Kylling, A., Systematic long-term comparison of spectral UV measurements and UVSPEC modeling results. Journal of Geophysical Research 102, D7, Mayer, B., Kylling, A., Madronich, S., Seckmeyer, G., Enhanced absorption of UV radiation due to multiple scattering in clouds: experimental evidence and theoretical explanations. Journal of Geophysical Research 103, D23, Mayer, B., Kylling, A., Technical note: The libradtran sofrware package for radiative transfer calculations description and examples of use. Atmospheric Chemistry and Physics 5, Myhre, G., Samset, B.H., Schulz, M., Balkanski, Y., Bauer, S., Berntsen, T.K., Bian, H., Bellouin, N., Chin, M., Diehl, T., Easter, R.C., Feichter, J., Ghan, S.J., Hauglustaine, D., Iversen, T., Kinne, S., Kirkevåg, A.,

101 101 Lamarque, J.-F., Lin, G., Liu, X., Lund, M.T., Luo, G., Ma, X., van Noije, T., Penner, J.E., Rasch, P.J., Ruiz, A., Seland, Ø., Skeie, R.B., Stier, P., Takemura, T., Tsigaridis, K., Wang, P., Wang, Z., Xu, L., Yu, H., Yu, F., Yoon, J.-H., Zhang, K., Zhang, H., Zhou, C., Radiative forcing of the direct aerosol effect from AeroCom Phase II simulations. Atmospheric Chemistry and Physics 13, Nakajima, T., Tanaka, M., Yamauchi, T., Retrieval of the optical properties of aerosols from aureole and extinction data. Applied Optics 22, Nakajima, T., Tanaka, M., Algorithms for radiative intensity calculations in moderately thick atmospheres using a truncation approximation. Journal of Quantitative Spectroscopy & Radiative Transfer 40, Nakajima, T., Tonna, G., Rao, R., Boi, P., Kaufman, Y.J., Holben, B.N., Use of sky brightness measurements from ground for remote sensing of particulate polydispersions. Applied Optics 35, Pierluissi. J.H., Peng, G.-S., New molecular transmission band models for LOWTRAN. Optical Engineering 24, Quaas, J., Ming, Y., Menon, S., Takemura, T., Wang, M., Penner, J.E., Gettelman, A., Lohmann, U., Bellouin, N., Boucher, O., Sayer, A.M., Thomas, G.E., McComiskey, A., Feingold, G., Hoose, C., Kristjánsson, J.E., Liu, X., Balkanski, Y., Donner, L.J., Ginoux, P.A., Stier, P., Grandey, B., Feichter, J., Sednev, I., Bauer, S.E., Koch, D., Grainger, R.G., Kirkevåg, A., Iversen, T., Seland, Ø., Easter, R., Ghan, S.J., Rasch, P.J., Morrison, H., Lamarque, J.-F., Iacono, M.J., Kinne, S., Schulz, M., Aerosol indirect effects general circulation model intercomparison and evaluation with satellite data. Atmospheric Chemistry and Physics 9, Remer, L.A., Kaufman, Y.J., Tanré, D., Mattoo, S., Chu, D.A., Martins, J.V., Li, R.R., Ichoku, C., Levy, R.C., Kleidman, R.G., Eck, T.F., Vermote, E., Holben, B.N., The MODIS aerosol algorithm, products and validation. Journal of the Atmospheric Sciences 62, 4,

102 102 Ricchiazzi, P., Yang, S., Gautier, C., Sowle, D., SBDART: A research and teaching software tool for plane-parallel radiative transfer in the Earth s atmosphere. Bulletin of the American Meteorological Society 79, Schmetz, J., Pili, P., Tjemkes, S., Just, D., Kerkmann, J., Rota, S., Ratier, A., An introduction to Meteosat Second Generation (MSG). Bulletin of the American Meteorological Society 83, Schmid, B., Michalsky, J.J., Slater, W., Barnard, J.C., Halthore, R.N., Liljegren, J.C., Holben, B.N., Eck, T.F., Livingston, J.M., Russell, P.B., Ingold, T., Slutsker, I., Comparison of columnar water-vapor measurements from solar transmittance methods. Applied Optics 40, Schulz, M., Textor, C., Kinne, S., Balkanski, Y., Bauer, S., Berntsen, T., Berglen, T., Boucher, O., Dentener, F., Grini, A., Guibert, S., Iversen, T., Koch, D., Kirkevåg, A., Liu, X., Montanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S., Seland, Ø., Stier, P., Takemura, T., Radiative forcing by aerosols as derived from the AeroCom present-day and pre-industrial simulations. Atmospheric Chemistry and Physics 6, Shettle, EP., Models of aerosols, clouds and precipitation for atmospheric propagation studies. AGARD Conference Proceedings No.454, Atmospheric Propagation in the UV, Visible, IR and MM-Wave region and Related System Aspects. Sinyuk, A., Dubovik, O., Holben, B.N., Eck, T.F., Breon, F.-M., Martonchik, J., Kahn, R.,Diner, D.J., Vermote, E.F., Roger, J.-C., Lapyonok, T., Slutsker, I., Simultaneous retrieval of aerosol and surface properties from a combination of AERONET and satellite data. Remote Sensing of Environment 107, Smirnov, A., Holben, B.N., Lyapustin, A., Slutsker, I., Eck, T.F., AERONET processing algorithms refinement. AERONET Workshop, May 10-14, 2004, El Arenosillo, Spain.

103 103 Stamnes, K., Tsay, S.-C., Wiscombe, W., Jayaweera, K., Numerically stable algorithm for discreteordinate-method radiative transfer in multiple scattering and emitting layered media. Applied Optics 27, Stamnes, K., Slusser, J., Bowen, M., Derivation of Total Ozone Abundance and Cloud Effects from Spectral Irradiance Measurements. Applied Optics 30, Tanré, D., Kaufman, Y.J., Herman, M., Mattoo, S., Remote sensing of aerosol properties over oceans using the MODIS/EOS spectral radiances. Journal of Geophysical Research 102, Textor, C., Schulz, M., Guibert, S., Kinne, S., Balkanski, Y., Bauer, S., Berntsen, T., Berglen, T., Boucher, O., Chin, M., Dentener, F., Diehl, T., Easter, R., Feichter, H., Fillmore, D., Ghan, S., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Horowitz, L., Huang, P., Isaksen, I., Iversen, T., Kloster, S., Koch, D., Kirkevåg, A., Kristjánsson, J.E., Krol, M., Lauer, A., Lamarque, J.F., Liu, X., Montanaro, V., Myhre, G., Penner, J., Pitari, G., Reddy, S., Seland, Ø., Stier, P., Takemura, T., Tie, X., Analysis and quantification of the diversities of aerosol life cycles within AeroCom. Atmospheric Chemistry and Physics 6, Textor, C., Schulz, M., Guibert, S., Kinne, S., Balkanski, Y., Bauer, S., Berntsen, T., Berglen, T., Boucher, O., Chin, M., Dentener, F., Diehl, T., Feichter, J., Fillmore, D., Ginoux, P., Gong, S., Grini, A., Hendricks, J., Horowitz, L., Huang, P., Isaksen, I.S.A., Iversen, T., Kloster, S., Koch, D., Kirkevåg, A., Kristjansson, J.E., Krol, M., Lauer, A., Lamarque, J.F., Liu, X., Montanaro, V., Myhre, G., Penner, J.E., Pitari, G., Reddy, S., Seland, Ø., Stier, P., Takemura, T., Tie, X., The effect of harmonized emissions on aerosol properties in global models an AeroCom experiment. Atmospheric Chemistry and Physics 7, Woods, T.N., Prinz, D.K., Rottmann, G.J., London, J., Crane, P.C., Cebula, R.P., Hilsenrath, E., Brueckner, G.E., Andrews, M.D., White, O.R., VanHoosier, M.E., Floyd, L.E., Herring, L.C., Knapp, B.G., Pankrantz, C.K., Reiser, P.A., Validation of the UARS solar ultraviolet irradiances: Comparison with the Atlas 1 and 2 measurements. Journal of Geophysical Research 101,

104 Aerosol optical properties and corresponding solar irradiances in the Mediterranean Abstract The variability of aerosol optical properties over the Mediterranean basin, in the ultraviolet (UV) and visible (VIS) spectral regions, is examined, using ground-based data from eight stations of the AErosol RObotic NEtwork (AERONET), climatological values from the AeroCom (Aerosol Comparisons between Observations and Models) project and data from the MODerate resolution Imaging Spectroradiometer (MODIS) on-board the Terra and Aqua satellites. The aerosol optical properties from the three datasets are used as input data in a radiative transfer model, in order to estimate the irradiance at the ground under cloudfree skies. The differences of the modeled (AeroCom) and satellite-derived (MODIS) data, from the groundbased measurements (AERONET) are studied. The MODIS aerosol optical depth climatology shows better agreement with AERONET data. The highest difference in the monthly average values is equal to 0.09 at 550nm, while the differences between the AERONET and the AeroCom climatologies reach 0.25 and 0.15 in the UV and VIS wavelengths respectively. As a result, the AERONET modeled VIS and UV irradiances are closer to MODIS, with the absolute differences in average values reaching 15 W/m 2 (6%) and 2W/m 2 (6%) respectively, while absolute differences with AeroCom irradiances can reach up to 30 (12%) and 4.5 W/m 2 (12%). The differences are higher in FORTH Crete, Nes Ziona, Sede Boker and Blida, which are influenced by desert dust aerosols. The separate effect on irradiances due to differences in aerosol optical depth and single scattering albedo from the three datasets has been also examined; according to results, the effect of aerosol optical depth is dominant. At this stage, the use of MODIS aerosol climatology would be recommended for calculations of UV and VIS irradiance in the Mediterranean basin. However, the AeroCom climatology, with some further improvements in the dust cycle, could be a valuable tool in future studies in this area, due to the provision of natural and manmade aerosol optical properties with higher spectral resolution. 1 Based on the publication: Nikitidou and Kazantzidis,

105 Introduction Aerosols emitted in the atmosphere through natural processes (e.g. dust, sea salt, volcano eruptions) and anthropogenic activities (e.g. fossil fuel, industries, transport) have a direct effect, by scattering and absorbing the incoming solar radiation, and indirect effect, by altering cloud properties, in the Earth s energy balance. The radiative forcing efficiency of aerosols holds one of the highest uncertainties (IPCC 2007) and ground-based measuring networks, global aerosol models and satellite instruments have been developed to estimate their optical properties, formation, transportation and role in the climate system. The AERONET is a worldwide ground network established by NASA for the study of atmospheric aerosols (Holben et al., 2001). It comprises over 100 stations which continuously provide data of aerosol optical properties, based on direct sun and sky radiance measurements by Cimel sun photometers (Holben et al., 1998, Dubovik and King, 2000, Dubovik et al., 2000, Dubovik et al., 2002, Gobbi et al., 2007). The AERONET data have been used extensively in studies related to the validation of satellite aerosol products retrievals (e.g. Chu et al., 2002, Torres et al., 2002, Kahn et al., 2005, Vidot et al., 2008, De Meij and Lelieveld., 2011), validation of other ground-based instruments aerosol properties (e.g. Gröbner and Meleti, 2004, Cheymol et al., 2009), the spatio-temporal variability of aerosols characteristics (e.g. Smirnov et al., 2002, Bäumer et al., 2008, Georgoulias and Kourtidis, 2011) and the calculation of aerosol direct radiative forcing (e.g. Garcia et al., 2006, Bergamo et al., 2008, Clerici and Mélin, 2008). The AeroCom project is an international effort aiming to a better understanding of the effect of aerosols in climate. Estimations from more than 14 global models were compared with each other and with satellite and ground data, in order to improve the aerosol modules in the global models and minimize the uncertainty of their radiative effect (Kinne et al., 2006). These data, representing the AeroCom ensemble model, are provided with a spatial resolution of 100x100 km (1 x1 ). An initial assessment of AeroCom aerosol properties was done by Kinne et al. (2006). In general, a quite successful agreement was found in the annual global average of aerosol optical depth, at 550 nm, with AeroCom values being at 0.11 to 0.14 and AERONET at However, differences in regional distribution and the contribution of dust and carbonaceous aerosols have been revealed (Textor et al, 2007), indicating that further improvement and

106 106 understanding of the aerosol processes, chemistry, transport and removal mechanisms is needed. An estimation of the aerosol direct radiative forcing, based on 9 AeroCom models, was provided by Schulz et al. (2006), giving a global annual value of Wm -2 at the top of the atmosphere. The dust transport and deposition was addressed in Huneeus et al. (2011), where 15 AeroCom models were compared with each other in the reproduction of the dust characteristics and its direct radiative effect. According to results, considerable differences were revealed among the models in the simulation of the dust cycle and, generally, the aerosol optical depth and the particle size were overestimated. In order to gain better spatial coverage, measuring systems in satellites have been used to observe the Earth s atmosphere, land and ocean. MODIS is the instrument onboard the NASA s Terra and Aqua satellites which has been providing data regarding the aerosol optical characteristics since February 2000 (Terra) and July 2002 (Aqua). The two satellites are in a sun-synchronous, near-polar, circular orbit and view the entire Earth s surface every 1-2 days (Remer et al., 2005). MODIS data have been used to study the seasonal variation of aerosols and discrimination of their different types, biomass burning aerosols and dust events (e.g. Arola et al., 2007, Santese et al., 2007), as well as calculate the aerosol direct radiative effect (Zhou et al., 2005). The new algorithm used for processing the acquired data, labeled Collection 5, (Levy et al., 2007a, 2007b) provides more accurate results when compared to AERONET. Remer et al. (2008) showed that the new algorithm improves the retrievals over land; over non-bright surfaces, the expected error in aerosol optical depth (AOD) over land is within ±( AOD). However, the reliability of data over bright surfaces (e.g. deserts) is still limited. Papadimas et al. (2009) compared the MODIS Collections 4 and 5 and AERONET data over the Mediterranean basin and stated that their correlation coefficient was increased from 0.66 to 0.76 and the overall differences were within the expected range of uncertainties of the two datasets. The area of this study, the Mediterranean basin, exhibits a large variety of aerosols, such as desert dust from North Africa and the Middle East, maritime aerosols from the Mediterranean Sea, as well as anthropogenic aerosols from the European coastal cities, resulting in significantly higher concentrations (from 2 to 10 times) than over the least polluted environments at northern latitudes (Lelieveld et al, 2002).

107 107 Several studies have been conducted for this area, using both ground-based and satellite retrieved datasets. Based on MODIS data, Papadimas et al. (2008) showed that there has been a 20% decrease in aerosol optical depth in the region, during the period , observed mainly in the western parts of the Iberian, Italian and Balkan peninsulas and the Southern Anatolian peninsula. Koukouli et al. (2010), revealed a decreasing trend in aerosol load between -2.5 and -4.1% per year over South Balkan/Eastern Mediterranean region. Based on the same dataset, Gkikas et al. (2009) studied the frequency and intensity of aerosol episodes, showing that the western and central Mediterranean areas are subject to strong aerosol events, associated with sea salt and biomass burning cases. Eastern Mediterranean regions are influenced, besides the local anthropogenic sources, by the desert dust aerosols and their transport, reaching high values in spring and summer (Hatzianastassiou et al., 2009, Querol et al., 2009, Gerasopoulos et al., 2011, Kalivitis et al., 2011). Similar results have been reported from the analysis of aerosol data from single ground-based sites, selected for this study. The Sede Boker station (31 o N, 35 o E), in the Negev desert of Israel, exhibits a high aerosol optical depth in spring due to desert dust and in late summer-early autumn attributed to anthropogenic aerosols. However, the radiative effect of mineral dust is considered as dominant (Derimian et al., 2006). The FORTH station (35 o N, 25 o E) in the island of Crete, Greece, presents a spring maximum due to dust transport and a winter minimum. During summertime, the transport of industrial and biomass burning aerosols is dominant (Fotiadi et al., 2006). This site is ideal for the study of transport and mixing of different aerosol types. Thessaloniki (41 o N, 23 o E), in northern Greece, is also a crossroad for a variety of aerosols, being additionally influenced by aerosols of industrial origin from the northeastern European countries (Balis et al., 2004). Kelektsoglou and Rapsomanikis (2011) revealed a summer maximum and a winter minimum, but reported high values in spring and autumn as well. More absorbing particles were found during autumn and winter, mainly due to emissions of aerosols from local pollution sources. A nine year ( ) period study of the aerosol optical depth in the UV region and the PM10 concentrations showed a decreasing tendency, but the levels remained high (Kazadzis et al., 2007). The station at the IMS- METU site (37 o N, 34 o E) in Erdemli, Turkey is ideal for the study of mineral dust characteristics, since it is affected by dust transport from both the Sahara and Middle East deserts. Sahara dust episodes occur more

108 108 frequently in spring, whereas dust from the Middle East deserts affects aerosol concentration at higher altitudes during summer and autumn (Kubilay et al., 2003). The station at Lecce University (40 o N, 18 o E), south Italy, is in a remote location, not influenced by industrial sources. It is generally characterized by clean continental aerosols and a mixing of water-soluble and more absorbing particles is evident (Barnaba et al., 2007). In this chapter, the differences in aerosol optical properties (optical depth, single scattering albedo and asymmetry factor) in the UV and VIS spectral regions over eight sites in the Mediterranean basin are examined, using data from the AERONET, the AeroCom climatology and the MODIS satellite instrument. A radiative transfer model is used to estimate the uncertainties in UV and VIS irradiance reaching the ground that arise from the use of the satellite and model-derived climatologies Data and Methodology The monthly climatological values of aerosol optical depth, single scattering albedo and asymmetry factor retrievals from 8 AERONET stations across the Mediterranean Sea were selected for this study (Figure 3.1), covering the time period

109 109 Figure 1.1: Location map of the 8 AERONET stations across the Mediterranean basin. These stations are Blida (37 o N, 3 o E), Carpentras (44 o N, 5 o E), Lecce University (40 o N, 18 o E), Thessaloniki (41 o N, 23 o E), FORTH Crete (35 o N, 25 o E), IMS-METU Erdemli (37 o N, 34 o E), Nes Ziona (32 o N, 35 o E) and Sede Boker (31 o N, 35 o E). More recent data were not used, since the AeroCom climatology derived from model runs averaged for 3-10 years (Kinne et al., 2006) and is considered to be representative for the time period (Pappas et al., 2011). AERONET level 2.0 data are used, which are pre- and post-field calibrated, automatically cloud cleared and manually inspected. More specifically, the instruments are calibrated before they are positioned in the field and after a measurement period. A cloud-screening algorithm is used by applying criteria for the short time period variability (1 minute) as well as the hourly and diurnal time period variations of AOD. Finally, data are visually inspected for abnormal data points. Blida Carpentras FORTH Crete IMS- METU Erdemli Lecce University Nes Ziona Sede Boker Thessaloniki Oct2003- Jun2008 [ Years] Feb2003- May2008 [ Years] Jan2003- Mar2008 [ Years] Nov1999- Jan2009 [ Years] Mar2003- Dec2008 [ Years] Feb2000- Mar2008 [ Years] Jan1996- Jan2008 [ Years] Sep2005- Jan2008 [ Years] Table 3.1: Time periods of available level 2.0 AERONET AOD data, used for each station.

110 110 The use of level 2.0 data and the different starting dates of operation at each station reduced the total amount of years from 1.5 (at Thessaloniki) to 9.2 (at Sede-Boker) (Table 3.1). For the theoretical calculations, the monthly mean values of the α and β Ångstrom coefficients, the single scattering albedo and the asymmetry factor were used as input data. For calculations of UV irradiance, the α and β Ångstrom coefficients were derived from the optical depths at 340 and 380 nm. AERONET does not provide the single scattering albedo and asymmetry factor values in the UV region; so, measurements at the lower available wavelength (440nm) were used also for UV calculations. Similarly, in the visible spectral region, the Angstrom coefficients were calculated from the optical depths at 440 and 675nm. The average values of the single scattering albedo and the asymmetry factor at those two wavelengths were used in model calculations. For UV and VIS irradiance calculations based on AeroCom climatology, the Ångstrom coefficients, derived from the optical depths at two UV (345 and 380nm) and VIS (440 and 665nm) wavelengths, were used. The average values of single scattering albedo and asymmetry factor at these wavelengths were used for the model calculations. The Ångstrom coefficients, single scattering albedo and asymmetry factor, that were used as input in the model calculations, refer to the monthly mean values for the time considered. From the MODIS dataset, the monthly climatological averages of Ångstrom exponent and aerosol optical depth at 550nm over each site for the period were taken into account. The daily averaged values, derived from the Terra and Aqua satellite estimations when they were available, were used to calculate the monthly climatological averages for the selected time period. Due to the limitation of MODIS operation in VIS and infrared wavelengths, the aerosol optical depth was extrapolated in the UV, using the Ångstrom formula, τ(λ 1 )=τ(λ 2 )*(λ 1 /λ 2 ) -α. Since MODIS dataset does not provide estimations about the single scattering albedo and the asymmetry factor, spectrally independent values (0.95 and 0.67 respectively) were used in model calculations of irradiance. These values were taken as representative of the Mediterranean basin and similar values are found by Bergamo et al., (2008). The estimation of the irradiance at the ground was done using the UVspec code from the LibRadtran package (Mayer and Kylling, 2005). Typical, for the mid-latitudes, vertical profiles of the basic atmospheric gases, aerosols, pressure and temperature were used. Aerosols were further described with the assigned

111 111 values for α and β Angstrom coefficients, single scattering albedo and asymmetry factor. The total ozone column was set to 320 DU. The surface reflectance was set to 0.03 in the UV and 0.10 in the VIS region. The value for the reflectance in the UV agrees with the ones given in Herman and Celarier, (1997) for the Mediterranean area, while the value in the VIS was chosen as a combination of the ones given in Bowker et al., (1985), for sand, water and concrete surfaces. Since the study focuses on differences between the modeled irradiances for the three datasets, the accuracy in the input reflectances doesn t affect the results. The purpose of this study was limited to the estimation of differences in cloud-free irradiance due to the use of different aerosol datasets. At each station and for the 15 th day of each month, the monthly climatological values of aerosol optical properties from AERONET, AeroCom and MODIS were taken into account to calculate the UV ( nm) and VIS ( nm) integrated irradiance from sunrise to sunset in 30 minute steps including the time of local noon Results and Discussion The monthly climatological averages of aerosol optical depth from AERONET, AeroCom and MODIS, all extrapolated at 550 nm, using the Ångstrom formula, over each station are presented in Figure 3.2. Due to the limited availability of MODIS data over Sede Boker, only results for AERONET and AeroCom are presented. Generally, MODIS data seem to be in better agreement with AERONET values over all other sites. At Blida, a very good agreement between them is revealed, while AeroCom values are significantly lower (down to 0.24) from April to November. At Lecce University and FORTH Crete, the aerosol optical depth is overestimated by AeroCom during most months of the year and the peak value of May is not evident in the other two datasets. Lower differences among the three datasets are revealed at Carpentras, Thessaloniki and IMS-METU Erdemli during some months. In contrast with the results from the North African site of Blida, the ground-based data are significantly lower than satellite and model estimations at the Middle East sites of Nes Ziona and Sede Boker.

112 112 Figure 3.2: Monthly average values of AOD at 550 nm from AERONET, AeroCom and MODIS over the selected sites. No MODIS data were available at Sede Boker station. Figure 3.3 presents the differences in AOD at 380 nm. MODIS AOD has been extrapolated in the UV using the Angström formula, τ(λ)=β*λ -α, where α and β are the Angström coefficients. The average

113 difference at each site is within ±0.03, relatively to the difference at 550nm. However, in certain months, the differences are significantly higher. 113 Figure 3.3: Monthly average values of AOD at 380 nm from AERONET, AeroCom and MODIS over the selected sites. No MODIS data were available at Sede Boker station.

114 114 In general, the highest differences from the AERONET data are revealed over the Middle East and North African sites and the island of Crete and, most probably, are relevant to the limited accuracy of MODIS aerosol products over bright surfaces and the imprecise simulation of the dust cycle in the AeroCom models. It should be noted that the AERONET data at the selected sites are not available during the exact time period of MODIS data or the period that AeroCom climatology is considered as representative. However, the presentation of these results highlights the uncertainties of the satellite and modeled climatologies over the Mediterranean basin and indicates the importance to examine their effect in calculations of UV and VIS irradiance reaching the ground. SSA Aerocom AERONET Stations 380 nm 420 nm 665 nm 440 nm 675 nm Blida Carpentras FORTH Crete IMS-METU Erdemli Lecce University Nes Ziona Sede Boker Thessaloniki Table 3.2: Annual average values of SSA from AERONET and AeroCom over the selected sites. Solar irradiance reaching the ground is also affected by the single scattering albedo and, less significantly, by the asymmetry factor of aerosols. The average annual values of the AERONET and AeroCom single scattering albedo over each station at similar wavelengths in the VIS spectral region and at 380 nm for AeroCom, are presented in Table 3.2. In all cases, the average overestimation of single scattering

115 albedo by the AeroCom climatology is around Similar results are revealed when comparing the AeroCom values in the UV region with the AERONET ones at 440nm. 115 Figure 3.4: Percentage (%) differences of monthly averages of UV irradiance calculations at the selected sites, when using the AeroCom dataset relatively to the AERONET ground-based measurements. With black symbols are the differences when both AOD and SSA are used from the AeroCom dataset, red and blue are the differences when only SSA and AOD is used, respectively.

116 116 The monthly mean percentage (%) differences of calculated average UV irradiance between the AeroCom climatology and the AERONET measurements are presented in Figure 3.4, for each station. The combined differences are those resulting from the use of both AOD and SSA from AeroCom dataset. The separate effect of each factor is investigated and the differences due to the use of only SSA or AOD are also presented in the figure. In general, the greatest monthly differences are within ±15% and correspond to almost ±4.5 W/m 2 in UV irradiance reaching the ground. During the April-September period, the UV irradiance derived from the AeroCom climatology is underestimated at most sites. The underestimation is higher (14%) for sites of the Eastern Mediterranean that lie close to desert regions (Sede Boker, Nes Ziona, FORTH Crete). For the same time period, less UV values (down to -7%) are calculated at Lecce University and IMS-METU Erdemli. At Thessaloniki and Carpentras, a quite good agreement (within 5 and 8%) is revealed for almost all months. On the contrary, the use of AeroCom climatology leads to the overestimation of UV irradiance up to 15% at another site that is affected dominantly from desert dust, the North African station of Blida. Significant overestimations (above 10%) can be observed also at specific sites and months, like February at IMS-METU Erdemli and Nes Ziona and November-December at Thessaloniki. Similar results were calculated for UV irradiance at local noon (Figure 3.5), but, due to the weaker effect of aerosols on irradiance at lower solar zenith angles, the overall agreement is within ±12%. The maximum percentage difference corresponds to ~ 7 W/m 2 in UV irradiance reaching the ground.

117 117 Figure 3.5: Percentage (%) differences of monthly UV irradiance calculations, at local noon, at the selected sites, when using the AeroCom dataset relatively to the AERONET ground-based measurements. Figure 3.6 : : Percentage (%) differences of monthly VIS irradiance calculations, at local noon, at the selected sites, when using the AeroCom dataset relatively to the AERONET ground-based measurements.

118 118 In general, as it can be observed in Figure 3.4, the differences in calculated irradiances are due to AOD values, while the effect of SSA is of secondary importance. However, some exceptions are revealed at some stations during specific months. For example, the high difference of 12% at Nes Ziona in February is almost entirely attributed to the difference in SSA values between the AeroCom and AERONET datasets. For the majority of months, the differences due to the use of SSA from AeroCom are around 2%, except in Blida, where the differences vary between 0 and 6%. In general, increased irradiance values are calculated by using the AeroCom SSA, since it is higher than the corresponding values from the AERONET database, as already seen in Table 3.2. The mean percentage differences between the modeled VIS irradiance, as derived from AeroCom and AERONET datasets are presented in Figure 3.7. Percentage differences up to ±12% are revealed which correspond to discrepancies in irradiance up to 30W/m 2. Similar calculations for local noon (Figure 3.6), reveal differences up to 10% (35W/m 2 ). As it can be observed in Figure 3.7, AOD is again the major factor responsible for the irradiance differences. Differences due to SSA range between 0 and 4% and are positive in their majority.

119 Figure 3.7: Percentage (%) differences of monthly averages of VIS irradiance calculations at the selected sites, when using the AeroCom dataset relatively to the AERONET ground-based measurements. With black symbols are the differences when both AOD and SSA are used from the AeroCom dataset, red and blue symbols are the differences when only SSA and AOD is used, respectively. 119

120 Figures 3.8 and 3.11 present the monthly mean percentage differences between the modeled UV and VIS irradiances, respectively, as derived from MODIS and AERONET datasets. 120 Figure 3.8: Percentage (%) differences of monthly averages of UV irradiance calculations at the selected sites, when using MODIS datasets relatively to the AERONET ground-based measurements. Upper panel shows the differences when the constant value of 0.95 is used for the SSA in the MODIS irradiance calculations while in the lower panel SSA values are used from the AERONET dataset.

121 121 Figure 3.9 : Percentage (%) differences of monthly UV irradiance calculations, at local noon, at the selected sites, when using MODIS datasets relatively to the AERONET ground-based measurements. Figure 3.10 : Percentage (%) differences of monthly VIS irradiance calculations, at local noon, at the selected sites, when using MODIS datasets relatively to the AERONET ground-based measurements.

122 122 Figure 3.11: Percentage (%) differences of monthly averages of VIS irradiance calculations at the selected sites, when using the MODIS datasets relatively to the AERONET ground-based measurements. Upper panel shows the differences when the constant value of 0.95 is used for the SSA in the MODIS irradiance calculations while in the lower panel SSA values are used from the AERONET dataset. In each of the two figures, the upper panel presents the differences that arise with the use of AOD from MODIS and the yearly constant value of 0.95 for SSA, while in the lower panel SSA from the AERONET dataset is used in the calculation of the modeled MODIS irradiances, so the differences are attributed directly to the MODIS AOD. The agreement between UV irradiance calculations with MODIS and AERONET climatologies, when compared with the AeroCom-AERONET differences, is significantly

123 123 better. During winter and spring, MODIS-derived UV irradiance is lower by ~3%, while the average difference is reduced to 0% for the rest of the year. Again, the highest differences during summertime are calculated over Blida (6%). The corresponding difference in average and local noon (Figure 3.9) UV irradiance reaching the ground is 2 and 3.3 W/m 2 respectively. If SSA from AERONET is used in the calculations of the modeled MODIS UV irradiances, then increased differences are revealed, since AERONET SSA, as given in Table 3.2, is usually lower than The use of the same SSA value as model input reveals the real effect of AOD differences between the two datasets in model calculated UV irradiance. In this case also, the effect of SSA is of secondary importance. At all sites, the MODIS-derived VIS irradiance is lower than the AERONET one during most months (Figure 3.11). However, the maximum underestimation in average values is 7% (15W/m 2 ). At local noon (Figure 3.10), differences up to 20W/m 2 (6%) are calculated. The highest differences are revealed over FORTH Crete and Nes Ziona stations during the winter and spring respectively. During summertime, when the effect of aerosols on solar irradiance is more pronounced due to decreased cloudiness, the overall difference ranges from -4 to +2%. Again, the use of AERONET SSA in the calculation of MODIS irradiances, increases the differences between the two datasets, but the increase is lower than in the UV, since aerosols affect less the irradiance at visible wavelengths than the ultraviolet ones Conclusions The aerosol optical properties from the AeroCom climatology and the MODIS satellite instrument were used to examine their differences from ground-based measurements and estimate the effect on cloudfree UV and VIS irradiance calculations over eight AERONET sites across the Mediterranean Basin. The absolute differences in aerosol optical depth climatologies from AeroCom and AERONET can reach up to 0.25 and 0.15, in the UV and VIS part of the spectrum respectively. These maximum differences occur in spring, where the aerosol loads are higher. The maximum absolute difference in aerosol optical depth from AERONET and MODIS at 550 nm, also occurs in spring, but is quite lower than the one from AeroCom,

124 reaching the value of During the rest of the year, the AERONET climatology is in better agreement with the MODIS than the AeroCom data. 124 The monthly mean values of aerosol optical properties from AERONET, AeroCom and MODIS were used to estimate the UV ( nm) and VIS ( nm) irradiance at the ground. Differences between AERONET and AeroCom in the UV can reach up to 7 (12%) and 4.5 W/m 2 (15%) for local noon and average values respectively. In the VIS spectral region, the corresponding values are 35 (10%) and 30 W/m 2 (12%). The largest irradiance differences are observed at the North African and Middle East stations of the Mediterranean coast (Nes Ziona, Sede Boker, Blida) and the island of Crete. The differences between the AERONET and the MODIS-derived VIS irradiances are lower, reaching 20 (6%) and 15 W/m 2 (7%) for the local noon and the average values respectively. The highest values correspond to the stations of FORTH Crete and Nes Ziona. On average, the difference in UV irradiance is between 0 and 3%. During winter and spring, MODIS-derived UV irradiance is lower by ~3%, while the average difference is reduced to 0% for the rest of the year. Again, the highest differences during summertime are calculated over Blida (6%). The corresponding difference in average and local noon UV irradiance reaching the ground is 2 and 3.3 W/m 2 respectively. In general, the pre-mentioned differences in calculated irradiances are mostly attributed to the AOD values, while the effect of SSA is of secondary importance. The MODIS and the AeroCom climatologies of aerosol optical properties present the highest differences from the AERONET ground-based measurements, over sites where the presence of desert dust aerosols is dominant. MODIS estimations are closer to the AERONET values, which indicate the need for further improvement of the dust cycle in the AeroCom models. A thorough comparison has been attempted among climatologies of aerosol optical properties in the Mediterranean basin, taken from AeroCom, MODIS and AERONET, and their effect on calculations of UV and VIS irradiance. According to the results, the satellite and modeled datasets still need improvements and the corresponding climatologies have to be further assessed. The MODIS climatology is in better agreement

125 125 with ground-based measurements and could be preferred for calculations of UV and VIS irradiance. However, with some further improvements, the AeroCom climatology could be a significantly valuable tool, since it provides estimations of aerosol optical properties with higher spectral resolution about natural and anthropogenic aerosols.

126 126 References Arola A, Lindfors A, Natunen A, Lehtinen KEJ A case study on biomass burning aerosols: effects on aerosol optical properties and surface radiation levels. Atmospheric Chemistry and Physics 7, Balis DS, Amiridis V, Zerefos C, Kazantzidis A, Kazadzis S, Bais AF, Meleti C, Papayannis A, Matthias V, Dier H Study of the effect of different type of aerosols on UV-B radiation from measurements during EARLINET. Atmospheric Chemistry and Physics 4, Barnaba F, Tafuro AM, De Tomasi F, Perrone MR Observed and simulated vertically resolved optical properties of continental aerosols over southeastern Italy: A closure study, Journal of Geophysical Research 112, D10203, doi: /2006jd Bäumer D, Rinke R, Vogel B Weekly periodicities of Aerosol Optical Thickness over Central Europe evidence of an anthropogenic direct aerosol effect. Atmospheric Chemistry and Physics 8, Bergamo A, Tafuro AM, Kinne S, De Tomasi F, Perrone MR Monthly-averaged anthropogenic aerosol direct radiative forcing over the Mediterranean based on AERONET aerosol properties. Atmospheric Chemistry and Physics 8, Cheymol A, Gonzalez Sotelino L, Lam KS, Kim J, Fioletov V, Siani AM, De Backer H Intercomparison of Aerosol Optical Depth from Brewer Ozone spectrophotometers and CIMEL sunphotometers measurements. Atmospheric Chemistry and Physics 9, Bowker DE, Davis RE, Myrik DL, Stacy K, Jones WT Spectral reflectance of natural targets for use in remote sensing studies. NASA Reference Publication Chu DA, Kaufman YJ, Ichoku C, Remer LA, Tanré D, Holben BN Validation of MODIS aerosol optical depth retrieval over land. Geophysical Research Letters 29, No. 12, doi: /2001gl Clerici M, Mélin F Aerosol direct radiative effect in the Po Valley region derived from AERONET measurements. Atmospheric Chemistry and Physics 8,

127 De Meij A, Lelieveld J Evaluating aerosol optical properties observed by ground-based and satellite remote sensing over the Mediterranean and the Middle East in Atmospheric Research 99, Derimian Y, Karnieli A, Kaufman YJ, Andreae MO, Anreae TW, Dubovik O, Maenhaut W, Koren I, Holben BN Dust and pollution aerosols over the Negev desert, Israel: Properties, transport, and radiative effect. Journal of Geophysical Research 111, D05205, doi: /2005jd Dubovik O, King MD A flexible inversion algorithm for retrieval of aerosol optical properties from Sun and sky radiance measurements. Journal of Geophysical Research 105, Dubovik O, Smirnov A, Holben BN, King MD, Kaufman YJ, Eck TF, Slutker I Accurancy assessments of aerosol optical properties retrieved from Aerosol Robotic Network (AERONET) Sun and sky radiance measurements. Journal of Geophysical Research 105, Dubovik O, Holben BN, Eck TF, Smirnov A, Kaufman YJ, King MD, Tanré D, Slutsker I Variability of absorption and optical properties of key aerosol types observed in worldwide locations. Journal of the Atmospheric Sciences 59, , doi: / (2002)059 <0590:VOAAOP>2.0.CO;2. Fotiadi A, Hatzianastassiou N, Drakakis E, Matsoukas C, Pavlakis KG, Hatzidimitriou D, Gerasopoulos E, Mihalopoulos N, Vardavas I Aerosol physical and optical properties in the Eastern Mediterranean Basin, Crete, from Aerosol Robotic Network data. Atmospheric Chemistry and Physics 6, García OE, Díaz AM, Expósito FJ, Díaz JP, Gröbner J, Fioletov VE Cloudless aerosol forcing efficiency in the UV region from AERONET and WOUDC databases. Geophysical Research Letters 33, L23803, doi: /2006gl Georgoulias AK, Kourtidis KA On the aerosol weekly cycle spatiotemporal variability over Europe, Atmospheric Chemistry and Physics 11, Gerasopoulos E, Amiridis V, Kazadzis S, Kokkalis P, Eleftheratos K, Andreae MO, Andreae TW, El- Askary H, Zerefos C.S Three-year ground measurements of aerosol optical depth over the Eastern Mediterranean: the urban environment of Athens. Atmospheric Chemistry and Physics 11,

128 Gkikas A, Hatzianastassiou N, Mihalopoulos N Aerosol events in the broader Mediterranean basin based on 7-year ( ) MODIS C005 data. Annales Geophysicae 27, Gobbi GP, Kaufman YJ, Koren I, Eck TF Classification of aerosol properties derived from AERONET direct sun data. Atmospheric Chemistry and Physics 7, Gröbner J, Meleti C Aerosol optical depth in the UVB and visible wavelength range from Brewer spectrophotometer direct irradiance measurements: Journal of Geophysical Research 109, D09202, doi: /2003jd Hatzianastassiou N, Gkikas A, Mihalopoulos N, Torres O, Katsoulis BD Natural versus anthropogenic aerosols in the eastern Mediterranean basin derived from multiyear TOMS and MODIS satellite data. Journal of Geophysical Research 114, D24202, doi: /2009jd Herman JR, Celarier EA Earth surface reflectivity climatology at nm from TOMS data. Journal of Geophysical Research 102, D23, Holben BN, Eck TF, Slutsker J, Tanré D, Buis JP, Setzer A, Vermote E, Reagan JA, Kaufman YJ, Nakajima T, Lavenu F, Jankowiak I, Smirnov A AERONET - A federated instrument network and data archive for aerosol characterization. Remote Sensing of Environment 66, Holben BN, Tanré D, Smirnov A, Eck TF, Slutsker I, Abuhassan N, Newcomb WW, Schafer J, Chatenet B, Lavenue F, Kaufman YJ, Vande Castle J, Setzer A, Markham B, Clark D, Frouin R, Halthore R, Karnieli A, ONeill NT, Pietras C, Pinker RT, Voss K, Zibordi G An emerging ground-based aerosol climatology: Aerosol Optical Depth from AERONET. Journal of Geophysical Research 106, Huneeus N, Schulz M, Balkanski M, Griesfeller J, Prospero J, Kinne S, Bauer S, Boucher O, Chin M, Dentener F, Diehl T, Easter R, Fillmore D, Ghan S, Ginoux P, Grini A, Horowitz L, Koch D, Krol MC, Landing W, Liu X, Mahowald N, Miller R, Morcrette J-J, Myhre G, Penner J, Perlwitz J, Stier P, Takemura T, Zender CS Global dust model intercomparison in AeroCom phase I. Atmospheric Chemistry and Physics 11,

129 Intergovernmental Panel on Climate Change (IPCC): Climate Change 2007, Working Group I Report: The physical science basis, Summary for Policymakers, Paris, February Kahn RA, Gaitley BJ, Martonchik JV, Diner DJ, Crean KA Multiangle Imaging Spectroradiometer (MISR) global aerosol optical depth validation based on 2 years of coincident Aerosol Robotic Network (AERONET) observations. Journal of Geophysical Research 110, D10S04, doi: /2004jd Kalivitis N, Bougiatioti A, Kouvarakis G, Mihalopoulos N Long term measurements of atmospheric aerosol optical properties in the Eastern Mediterranean. Atmospheric Research 102, Kazadzis S, Bais A, Amiridis V, Balis D, Meleti C, Kouremeti N, Zerefos CS, Rapsomanikis S, Petrakakis M, Kelesis A, Tzoumaka P, Kelektsoglou, K Nine years of UV aerosol optical depth measurements at Thessaloniki, Greece. Atmospheric Chemistry and Physics 7, Kelektsoglou K, Rapsomanikis S AERONET observations of direct and indirect aerosol effects over a South European conurbation. International Journal of Remote Sensing 32, 10, Kinne S, Schulz M, Textor C, Guibert S, Balkanski Y, Bauer SE, Berntsen T, Berglen TF, Boucher O, Chin M, Collins W, Dentener F, Diehl T, Easter R, Feichter J, Fillmore D, Ghan S, Ginoux P, Gong S, Grini A, Hendricks J, Herzog M, Horowitz L, Isaksen I, Iversen T, Kirkevåg A, Kloster S, Koch D, Kristjansson JE, Krol M, Lauer A, Lamarque JF, Lesins G, Liu X, Lohmann U, Montanaro V, Myhre G, Penner J, Pitari G, Reddy S, Seland Ø, Stier P, Takemura T, Tie X An AeroCom initial assessment optical properties in aerosol component modules of global models. Atmospheric Chemistry and Physics 6, Koukouli ME, Kazadzis S, Amiridis V, Ichokou C, Balis DS, Bais AF Signs of a negative trend in the MODIS aerosol optical depth over the Southern Balkans. Atmospheric Environment 44, Kubilay N, Cokacar T, Oguz T Optical properties of mineral dust outbreaks over the northeastern Mediterranean, Journal of Geophysical Research 108, D21, 4666, doi: /2003jd

130 130 Levy RC, Remer LA, Dubovik O. 2007a. Global aerosol optical properties and application 15 to Moderate Resolution Imaging Spectroradiometer aerosol retrieval over land. Journal of Geophysical Research 112, D13210, doi: /2006jd Lelieveld J, Berresheim H, Borrmann S, Crutzen PJ, Dentener FJ, Fischer H, Feichter J, Flatau PJ, Heland J, Holzinger R, Korrmann R, Lawrence MG, Levin Z, Markowicz KM, Mihalopoulos N, Minikin A, Ramanathan V, de Reus M, Roelofs GJ, Scheeren HA, Sciare J, Schlager H, Schultz M, Siegmund P, Steil B, Stephanou EG, Stier P, Traub M, Warneke C, Williams J, Ziereis H Global air pollution crossroads over the Mediterranean. Science 298, , doi: /science Levy RC, Remer LA, Mattoo S, Vermote EF, Kaufman YJ. 2007b. Second-generation operational algorithm: Retrieval of aerosol properties over land from inversion of moderate resolution imaging spectroradiometer spectral reflectance. Journal of Geophysical Research 112, D13211, 20, doi: /2006jd Mayer B, Kylling A Technical Note: The libradtran software package for radiative transfer calculations: Description and examples of use. Atmospheric Chemistry and Physics 5, Nikitidou, E., Kazantzidis, A., On the differences of ultraviolet and visible irradiance calculations in the Mediterranean basin due to model- and satellite-derived climatologies of aerosol optical properties. International Journal of Climatology, doi: /joc Papadimas CD, Hatzianastassiou N, Mihalopoulos N, Querol X, Vardavas I Spatial and temporal variability in aerosol properties over the Mediterranean basin based on 6-year ( ) MODIS data. Journal of Geophysical Research 113, D11205, doi: /2007jd Papadimas CD, Hatzianastassiou N, Mihalopoulos N, Kanakidou M, Katsoulis BD, Vardavas I Assessment of the MODIS Collections C005 and C004 aerosol optical depth products over the Mediterranean basin. Atmospheric Chemistry and Physics 9, Pappas V, Hatzianastassiou N, Papadimas C, Kinne S Evaluation of global AEROCOM aerosol optical properties against satellite MODIS aerosol products. Geophysical Reaserch Abstracts 13, EGU , EGU General Assembly 2011.

131 131 Querol X, Alastuey A, Pey J, Cusack M, Pérez N, Mihalopoulos N, Theodosi C, Gerasopoulos E, Kubilay N, Koçak M Variability in regional background aerosols within the Mediterranean. Atmospheric Chemistry and Physics 9, Remer LA, Kaufman YJ, Tanré D, Mattoo S, Chu DA, Martins JV, Li RR, Ichoku C, Levy RC, Kleidman RG, Eck TF, Vermote E, Holben BN The MODIS aerosol algorithm, products and validation. Journal of the Atmospheric Sciences 62, 4, Remer LA, Kleidman RG, Levy RC, Kaufman YJ, Tanré D, Mattoo S, Martins JV, Ichoku C, Koren I, Yu H, Holben BN Global aerosol climatology from the MODIS satellite sensors. Journal of Geophysical Research 113, D14S07, doi: /2007jd Santese M, De Tomasi F, Perrone MR Moderate Resolution Imaging Spectroradiometer (MODIS) and Aerosol Robotic Network (AERONET) retrievals during dust outbreaks over the Mediterranean. Journal of Geophysical Research 112, D18201, doi: /2007jd Schulz M, Textor C, Kinne S, Balkanski Y, Bauer S, Berntsen T, Berglen T, Boucher O, Dentener F, Guibert S, Isaksen ISA, Iversen T, Koch D, Kirkevåg A, Liu X, Montanaro V, Myhre G, Penner JE, Pitari G, Reddy S, Seland Ø, Stier P, Takemura T Radiative forcing by aerosols as derived from the AeroCom present-day and pre-industrial simulations. Atmospheric Chemistry and Physics 6, Smirnov A, Holben BN, Eck TF, Slutsker I, Chatenet B, Pinker RT Diurnal variability of aerosol optical depth observed at AERONET (Aerosol Robotic Network) sites. Geophysical Research Letters 29, No , doi: /2002gl Textor C, Schulz M, Guibert S, Kinne S, Balkanski Y, Bauer S, Berntsen T, Berglen T, Boucher O, Chin M, Dentener F, Diehl T, Feichter J, Fillmore D, Ginoux P, Gong S, Grini A, Hendricks J, Horowitz L, Huang P, Isaksen ISA, Iversen T, Kloster S, Koch D, Kirkevåg A, Kristjansson JE, Krol M, Lauer A, Lamarque JF, Liu X, Montanaro V, Myhre G, Penner JE, Pitari G, Reddy S, Seland Ø, Stier P, Takemura T, Tie X The effect of harmonized emissions on aerosol properties in global models an AeroCom experiment. Atmospheric Chemistry and Physics 7,

132 132 Torres O, Bhartia PK, Herman JR, Sinyuk A, Ginoux P, Holben B A Long-Term Record of Aerosol Optical Depth from TOMS Observations and Comparison to AERONET Measurements. Journal of the Atmospheric Sciences 59, , doi: / (2002)059<0398:ALTROA>2.0.CO;2 Vidot J, Santer R, Aznay O Evaluation of the MERIS aerosol product over land with AERONET. Atmospheric Chemistry and Physics 8, Zhou M, Yu H, Dickinson RE, Dubovik O, Holben BN A normalized description of the direct effect of key aerosol types on solar radiation as estimated from Aerosol Robotic Network aerosols and Moderate Resolution Imaging Spectroradiometer albedos. Journal of Geophysical Research 110, D19202, doi: /2005jd

133 Aerosol optical properties and radiative effects in the UV Abstract The measurements of aerosol optical depth, total ozone and UV irradiance from a Brewer spectrophotometer located at Uccle, Belgium, were used to estimate, for the first time at a typical site in Western Europe, the aerosol radiative forcing efficiency (the forcing performed per unit of aerosol optical depth). The study was performed at selected solar zenith angles during the period July 2006-May In the nm spectral region, the highest values were revealed at 30 o (-6.9 ± 0.9 W/m 2 ), while at 60 o the RFE was almost 2.5 times lower (-2.7 ±0.1 W/m 2 ). In the UV-B region ( nm), the RFE value at 60 o ( ±0.005 W/m 2 ) was 5 times lower than the corresponding value at 30 o (-0.35 ±0.04 W/m 2 ). Extending previous studies for the estimation of aerosol single scattering albedo in UV-A wavelengths down to 340nm, an attempt was made, taking advantage of the Brewer measurements, to provide estimates at low UV-A wavelengths and in the UV-B region. The estimated monthly averages of the Brewer single scattering albedo at 320 nm are in very close agreement (within ±0.01) with measurements at 440nm from a collocated CIMEL sunphotometer. Due to increased measurement uncertainties and the effect of ozone absorption, large differences between the two instruments were found at nm. For the rest of wavelengths, average differences up to 0.03 were revealed Introduction Aerosols are emitted in the atmosphere via natural (e.g. desert dust, sea salt, volcanic ash) and anthropogenic processes (e.g. transportation, industrial activities, fossil fuel burning). Aerosols influence the Earth s energy budget in a direct way, by scattering and absorbing the incoming solar radiation and in an indirect way, by changing the radiative properties and lifetime of clouds. Due to their complex effect in atmospheric processes, aerosols are one of the atmospheric constituents with the largest uncertainty regarding their radiative forcing (IPCC, 2007). 2 Based on the publication: Nikitidou et al.,

134 134 The UV radiation reaching the surface of the Earth is of extreme importance as its variations have a direct impact on living organisms and atmospheric chemistry (McKenzie et al., 2011, van Dijk et al., 2012). Under cloud-free skies, ozone and aerosols are the main atmospheric constituents that influence the transfer of UV radiation in the atmosphere. Since the first decade of the 21 st century, the stratospheric ozone recovery should lead to decreased UV irradiance at the ground (Tourpali et al., 2009, Bais et al., 2011). However, the different trends of aerosols in different places on Earth (Streets et al., 2009, Zhang and Reid, 2010, Yoon et al., 2011) affect downwelling UV irradiance (Zerefos et al, 2009, den Outer et al., 2010). In Europe, strong negative trends appear over most regions (Zhao et al., 2008, Karnieli et al., 2009, Koukouli et. al., 2010). Additionally, the aerosol optical depth (AOD) in Europe presents a weekly variability, having higher values during the weekdays due to increased transport and anthropogenic activities (Baumer et al., 2008, Xia et al., 2008, Georgoulias et al., 2011), while the seasonal variability shows that the lowest aerosol load is presented during wintertime (Koelemeijer et al., 2006, Kazadzis et al., 2007). Therefore, the optical depth and the radiative properties of aerosols in the UV spectral region are essential to be well established. In recent studies, the retrievals of AOD at UV wavelengths, based on measurements by Brewer spectrophotometers, were successfully compared with measured data from other instruments (Meleti and Cappellani, 2000, Marenco et al., 2002, Gröbner and Meleti, 2004, Sellitto et al., 2006, Cachorro et al., 2009, de Bock et al., 2010, Wilawan et al., 2012). Moreover, efforts have been made to estimate the aerosol radiative forcing efficiency (RFE, the forcing performed per unit of aerosol optical depth) in the UV range. Garcίa et al., 2006, showed that the aerosol RFE in surface UV irradiance ( nm) ranges, under cloudfree conditions, from -0.36±0.03 W/m 2 at Ispra, Italy, to -0.16±0.23 W/m 2 per unit of AOD at 340nm at two Canadian sites. In Western Mediterranean basin, the diurnal RFE in the nm spectral region (calculated by integrating the instantaneous forcing efficiencies from sunrise to sunset) was found to be ±0.45 W/m 2 and -3.22±0.61 W/m 2 per unit AOD at 380 nm, for the stations of Armilla and Sabinas respectively (Díaz et al., 2007). Kazadzis et al. (2009) found a mean reduction of irradiance in the wavelength range nm of 15.2% per unit of AOD slant column increase at Thessaloniki, Greece. Antón et al. (2011) calculated the erythemal forcing efficiency of aerosols in Granada, Spain, and reported values of -62 and -26 mw/m 2 per unit of AOD at 380 nm for solar zenith angles of 20 and 50 respectively,

135 135 showing the influence of solar zenith angle (sza). During an extreme desert dust event at El Arenosillo, Spain, Antón et al. (2012) reported a 50% decrease in the UV at 320 nm relative to days with low aerosol loads. The RFE values in the UV-B during that event were calculated at W/m 2 (for sza=30 ) and W/m 2 (for sza=50 ) per unit of AOD at 440 nm. Apart from AOD, another key parameter in the study of aerosol effect on radiative transfer is the single scattering albedo (SSA), which determines the scattering and absorbing efficiency of aerosols. Variations of SSA values could change the aerosol forcing from negative to positive, depending also on the altitude of aerosols and the surface albedo (Hansen et al., 1997). Measurements and modeled irradiances were used to derive the aerosol SSA in the UV region (Kylling et al., 1998, Kazantzidis et al., 2001, Petters et al., 2003, Lee et al., 2007, Corr et al., 2009, Ialongo et al., 2010). The uncertainty in the estimated effective SSA values depends on the radiation quantities used (global or direct-to-diffuse) and the aerosol load and increases in cases of low AOD values (Bais et al., 2005). Other uncertainties, related to the instrument measurements and the accuracy of input parameters (ozone and temperature profiles, AOD, surface albedo, asymmetry parameter) in theoretical models should also be taken into account in the estimated uncertainty of the retrieved effective SSA values (Kazantzidis et al., 2001, Kazantzidis et al., 2005, Cordero et al., 2007, Kazadzis et al., 2010). In this chapter, the aerosol RFE in the UV range was estimated for the first time at a typical West European site, Uccle, Belgium for the time period July May The retrieved AOD values at five wavelengths (306.3, 310.1, 313.5, and nm) were compared with the extrapolated ones at the same wavelength from the collocated CIMEL sunphotometer. The effect of sza was examined on aerosol RFE in the nm and nm spectral regions. Additionally, the aerosol effective SSA at the same wavelengths was estimated with the synergistic use of model calculations and measurements of global irradiance and compared with CIMEL data Data and Location The measurements used in this study were performed at the Royal Meteorological Institute at Uccle (50 48 N, 4 21 E), Belgium. Uccle is a residential suburb in Brussels and it is located at about 100 km from

136 136 the North Sea. The station is 100 m above mean sea level and is influenced, depending on the meteorological conditions, by urban continental and sea salt aerosols. In a previous study, Cheymol and De Backer (2003) used the Langley Plot Method to retrieve AOD values from the Brewer#016 single spectrophotometer s direct sun observations at five wavelengths (306.3, 310.1, 313.5, and nm) for the time period of Those retrieved AOD values showed a seasonal variation with values at nm around 0.4 and 0.9 in winter and summer respectively. Additionally, an annual trend of -2.46±0.37%/year at nm was found in AOD values for the period The comparison of the retrieved AOD at nm with those of a collocated CIMEL sunphotometer at 340 nm (extrapolated to nm), provided a correlation coefficient of 0.96 (Cheymol et al., 2009). De Bock et al. (2010) applied the Langley Plot Method on the double spectrophotometer s Brewer#178 measurements between 335 and 345 nm and convoluted them with the band pass function of the CIMEL filter at 340 nm, in order to derive the AOD at 340 nm and compare with the CIMEL values. The correlation coefficient, slope and intercept are 0.96, and respectively. The uncertainty of those Brewer AOD measurements was estimated to within ±0.08. The AOD values, used in this study, were retrieved by applying the Langley Plot Method to the direct sun measurements of the Brewer#178 instrument at five wavelengths (306.3, 310.1, 313.5, and 320.1nm). Total ozone measurements were provided by the same instrument along with UV measurements in the range nm. All data were obtained from July 2006 to May The instrument was recalibrated against the travelling standard Brewer instrument #017 in 2003, 2006 and 2008 (see The stability of ozone data has been continuously checked against the co-located instruments Brewer #016 and Dobson #40 (De Backer, 2009). The UV measurements were corrected for the cosine error, based on the methodology described in Lakkala et al. (2008). The uncertainty in global irradiance measurements is estimated to within ±7% (Gröbner et al., 2006). The variability of measurements from a collocated pyranometer was taken into account for the selection of the cloud-free measurements (Vasaras et al. 2001). AOD and total ozone measurements were linearly interpolated at the exact time of the UV measurement, since no synchronous measurements of all

137 137 three parameters can be measured by a Brewer spectrophotometer. The solar zenith angle (sza) at Uccle was calculated for every measurement at each wavelength. Data from a collocated CIMEL sunphotometer were also used. The CIMEL sunphotometer is part of the AERONET network (Holben et al., 1998, 2001, and performs direct sun measurements at 8 wavelengths (340, 380, 440, 500, 670, 870, 940 and 1020 nm) and sky radiance measurements at 4 wavelengths (440, 670, 870 and 1020 nm), providing the total column aerosol products through measurement of the solar extinction. CIMEL AOD level 2.0 and SSA level 1.5 data at 340 and 440 nm respectively were used for comparisons with the results from the Brewer measurements. The CIMEL SSA level 2.0 data were not used, due to their very limited availability (less than 20 common measurements during four years), so the level 1.5 data were chosen due to their higher availability for the considering time period. For the comparison of AOD values of the two instruments at the same wavelengths, the CIMEL α Angström exponent from AOD values at 340 and 440 nm was used for extrapolation to Brewer AOD measurements. In order to diminish further the possibility of cloud contamination and changing aerosol conditions, only quasi-simultaneous data from Brewer and CIMEL were chosen (max. time difference 3 min) Methodology The aerosol radiative forcing (RF) represents the difference between the irradiance received in a situation with aerosol present and one with aerosol-free conditions. The aerosol RF was calculated in the nm wavelength range, based on the Brewer measurements in the specified range and theoretical calculations under aerosol-free conditions with the uvspec radiative transfer model, part of the libradtran software package (Mayer and Kylling, 2005, The basic atmospheric gases, pressure and temperature profiles, typical for mid-latitudes, were used, while the ozone absorption cross section derived by Bass and Paur (1985). The surface albedo was set to 0.03, based on the climatological value provided by Herman and Celialer (1997). The discrete ordinate algorithm was used as the solver of the radiative transfer equation with the sphericity of the Earth taken into account. All modeled spectra were convoluted with the instrument s slit function. The global irradiance at the station altitude was estimated as a

138 138 function of sza and ozone in the wavelength range nm with the Brewer measuring step (0.5 nm). The integral in this wavelength range was calculated and a correction for the Earth-Sun distance was applied. The percentage RF (%) in the nm range was estimated as: I measured I mod eled RF (%) *100, (4.1) I mod eled where I measured is the integral of the Brewer UV measurements in the nm range and I modeled is the integral of the corresponding modeled irradiances for the same ozone and sza, under aerosol-free conditions. The radiative forcing efficiency (RFE) is the forcing performed per unit of aerosol optical depth. Using the RFE, the forcing no longer depends on the aerosol load but on the aerosol type and characteristics. The RFE is given by the slope of the linear regression between the RF and AOD values, RF (%) RFE (%). (4.2) AOD In order to estimate the aerosol effective SSA, another set of theoretical calculations was performed with uvspec adding aerosol parameters to the model input file. The aerosols profile was described according to Shettle et al. (1989). The asymmetry factor was set at 0.7 (constant in the UV spectrum) respectively, which could be considered as a typical value according to the collocated CIMEL measurements. Since AOD measurements were available at the wavelengths of 306.3, 310.1, 313.5, and nm, a Look-Up- Table (LUT) was calculated including the global irradiance, sza, total ozone, AOD and SSA. Each UV measurement was compared with the corresponding modeled irradiance at the same sza, ozone and AOD, as derived from the interpolated measurements. Only the SSA values for which the difference between the measured and modeled irradiance was within ±1% were accepted. The number of retrieved effective SSA values that satisfy this condition depends on the aerosol load; the lower the AOD, the higher the number of the retrieved effective SSA values and therefore the lower the accuracy in the effective SSA estimation (Bais et al., 2005).

139 Results Comparison with the CIMEL AOD The AOD values at nm, retrieved from the Brewer#178 measurements with the Langley Plot Method, were compared with AOD measurements from the CIMEL sunphotometer. The level 2.0 AOD data at 340 nm from CIMEL were extrapolated at nm using the Ångstrom formula, τ(λ 1 )=τ(λ 2 )*(λ 1 /λ 2 ) -α, where α is the AERONET Ångstrom coefficient derived from the AOD values at 340 and 440 nm. The comparison was made among data with a time difference of less than 3 minutes, in order to minimize the possibility of cloud contamination or aerosol changing conditions. The AODs are plotted in Figure 4.1. The values retrieved by the two instruments are in very good agreement, with the slope, intercept and r-square of the linear regression being equal to , and respectively. Figure 4.1: The aerosol optical depth (AOD) values at nm from CIMEL and Brewer instruments at Uccle. The slope, intercept and r-square are , and respectively Aerosol Radiative Forcing As previously mentioned, the RFE is the forcing performed per unit of aerosol optical depth. It was calculated from the integrals of the Brewer measurements at and nm and the modeled irradiances under aerosol-free conditions. The RFE values for the two spectral regions per unit of AOD at 320.1nm for the selected ranges of sza are presented in Table 4.1.

140 140 SZA Spectral Region 30 ±2 40 ±2 50 ±2 60 ± nm ± 0.9 W/m ±0.3 W/m ±0.1 W/m ±0.1 W/m nm ±0.04 W/m ±0.01 W/m ±0.005 W/m ±0.005 W/m 2 Table 4.3: The aerosol radiative forcing efficiency (RFE) at the selected solar zenith angles (sza) for two spectral regions. As expected, the RFE increases for decreasing sza (e.g. Di Biagio et al., 2010, Anton et al., 2011). In our study the higher value in the nm spectral region (-6.9 ± 0.9 W/m 2 ), was revealed at 30 o while at 60 o the RFE is almost 2.5 times lower (-2.7 ±0.1 W/m 2 ). Due to the different photon paths in the UV-B region ( nm), the RFE value at 60 o ( ±0.005 W/m 2 ) was 5 times lower than the corresponding value at 30 o (-0.35 ±0.04 W/m 2 ). Unfortunately, the results are not directly comparable with those presented in previous studies because all of them refer to RFE values in different spectral regions and AOD wavelengths. The homogenization of results in specific wavelength ranges and AOD values would be helpful for the detection of the effect of the different aerosol optical characteristics on RFE. In Figure 4.2, the RF(%) at nm, is presented for selected ranges of sza. For the highest AOD values, the RF(%) ranges between -15% (at 30 o ) and -35% (at 60 o ) depending on the sza of the measurement. The average difference between modeled and experimental values for clean conditions is estimated between -6.2% and +7.3%, depending on the sza. For the sza ranges of 30 ± 2, 40 ± 2, 50 ± 2 and 60 ± 2 the RFE (%) was ± 3.8%, ± 1.3%, -24 ± 0.8% and ± 1.1% respectively per unit of AOD at 320.1nm. Although in the case of the 30, the limited number of measurements does not lead to safe conclusions, it can be seen that the RFE (%) presents a short decrease as the sza increases. This could be probably attributed to the longer atmospheric paths for higher sza. The radiation received by the surface decreases in large sza due to the increase of Rayleigh scattering, especially in UV wavelengths, and becomes less sensitive to aerosols. This results in smaller slopes and, therefore, smaller RFE values. However, the observed differences are within the measurement uncertainties.

141 141 Figure 4.2: The aerosol radiative forcing (%) at Uccle in the nm wavelength range versus the Brewer AOD at nm for the selected solar zenith angles (30 ± 2, 40 ± 2, 50 ± 2 and 60 ± 2 ). It can be observed that for zero AOD values, the RF(%) is not equal to zero, as it would be expected if the aerosol-free modeled and measured irradiances were in complete agreement. The uncertainties related to the calculation of RF are linked with those mentioned in section 4.3. The effect of aerosols in the UV-B spectral region part ( nm) was also investigated (Figure 4.3). The RFE(%), as a function of the Brewer AOD at nm, was ± 2.8%, ± 1.2%, ± 0.7% and ± 1.1% around 30, 40, 50 and 60 respectively. In all cases, the calculated RFE(%) values for the UV-B region is lower by 2 3% relatively to the values for the nm spectral region.

142 142 Figure 4.3: The aerosol radiative forcing (%) at Uccle in the nm wavelength range versus the Brewer AOD at nm for the selected solar zenith angles (30 ± 2, 40 ± 2, 50 ± 2 and 60 ± 2 ). Figure 4.4: The aerosol radiative forcing (%) at Uccle in the nm wavelength range versus the Brewer AOD at nm, for CIMEL SSA values lower (left panel) and higher (right panel) than 0.9 respectively.

143 143 Figure 4.4, presents the RF(%) as a function of the Brewer AOD at nm for SSA measurements below and above 0.9. The SSA values from the CIMEL data at 440 nm were matched with the Brewer measurements, provided that the time difference between them was less than 30 minutes. The RFE(%) values, given by the slope of the linear regression, is ± 0.7% and ± 1.7% per unit of AOD at 320.1nm, for SSA values higher and lower than 0.9 respectively. So, in accordance with previous studies, the more absorbing aerosols lead to much higher RFE values. Di Biagio et al (2009) showed the decrease of aerosol RFE with increasing SSA values at the island of Lampedusa (central Mediterranean). Anton et al., (2011) classified their data according to the Angström exponent (from <0.5 for desert dust aerosols to >1.5 for urban and industrial aerosols) and the sza and revealed similar results for the erythemal dose rate at Granada, Spain. In our study, there was no discrimination of the data based on the sza of the measurement. As it was mentioned before, RFE(%) changes by only a few percent for the selected sza. Additionally, the Uccle site is not affected by desert dust. However, the calculated RFE(%) per unit of AOD at 320.1nm was higher by 14% for SSA measurements below 0.9. This difference could be safely attributed to the scattering/absorption efficiency of aerosols. Therefore, besides size, the strong absorption of urban and/or industrial aerosols affects significantly the RFE in UV at a typical West European site like Uccle Single Scattering Albedo Retrieval The effective SSA values were estimated at the 5 closest wavelengths (306.5, 310.0, 313.5, and nm) where Brewer AOD data were available. The effective SSA was retrieved by comparing the Brewer UV irradiance measurement at each wavelength with the corresponding modelled irradiances (using the total ozone, AOD and sza as model input parameters). In previous studies, this method had been applied and validated with other measurements in visible and UV-A wavelengths down to 340 nm. In this study, it was attempted to extend and validate the method at low UV-A wavelengths and in the UV-B region taking advantage of the Brewer AOD measurements at these wavelengths. As mentioned earlier, the effective SSA values were accepted when the difference between modelled and measured irradiance values is smaller than ±1%, since the uncertainty of the retrieval method depends on the aerosol load. In cases with low AOD, the effect on UV irradiance is diminished, the uncertainty of the method is higher and usually more than one

144 144 effective SSA value corresponds to difference within ±1% between modelled and measure irradiances. Bais et al. (2005) reported that for AOD=0.4 (at 340nm), the uncertainly in estimating the effective SSA from measurements of global irradiance is within ±0.08 and ±0.05 for sza of 30 o and 60 o respectively. However, the uncertainty is more than double for AOD=0.2. This can be seen in Figure 4.5, where four individual days are presented with different aerosol loads at nm. The Brewer and the CIMEL measurements (extended at the same wavelength with the use of the Angström exponent) during two days (27/3/2007 and 27/9/2009) described by high AOD, range from 0.35 to 0.7 and from 0.4 to 0.75 respectively (Figure 4.5, left panel). The application of the method retrieved only one effective SSA value for all measurements of those days and, in many cases, a very good agreement (within ±0.01) was revealed. However, during 30/4/2007 and 17/4/2010 (Figure 4.5, right panel), low AOD values (below 0.3 and 0.25 respectively) were observed. This resulted in low accuracy in the retrieval of the effective SSA values, providing up to 4 possible effective SSA values for each measurement.

145 145 Figure 4.5: Estimations of the effective single scattering albedo (SSA) at 320 nm from the Brewer method compared to CIMEL SSA values at 440 nm and the corresponding AODs from both instruments at Uccle. Due to the method limitation for low AOD values, only measurements with AOD 0.4 were taken into account in order to calculate the monthly average values of effective SSA. Moreover, the maximum difference in the AOD from Brewer and CIMEL was set at 5%. These two limitations ensure that close AOD values were taken into account and the differences that may emerge will be mainly due to the different methodologies for the SSA retrieval. Due to those limitations, it was not possible to use CIMEL SSA data of level 2.0, since less than 20 common measurements exist for the whole time-period.

146 146 Figure 4.6: The monthly averages of effective aerosol single scattering albedo (SSA) at Uccle from Brewer#178 at five wavelengths (306.5, 310, 313.5, and 320 nm) and CIMEL SSA at 440 nm with their standard deviations for the selected time period. The monthly averages of effective SSA from Brewer spectrophotometer at 306.5, 310.0, 313.5, and nm and CIMEL at 440 nm with their standard deviations, derived from at least 20 close measurements (within ±30 minutes) of the two instruments during each month, are presented in Figure 4.6. The adequate number of measurements was available only for March, April May and September. The estimated monthly values of Brewer effective SSA at 320 nm were in better agreement (within ±0.01 for monthly averages) with the CIMEL data. The estimated effective SSA averages ranged from 0.91 in March to 0.96 in May. At this wavelength, the calculated standard deviations were similar with those from the CIMEL. At nm, the Brewer-derived effective SSA was higher by 0.03 on average. At and nm, the monthly effective SSA averages from the Brewer instrument were lower by , compared to the values at 320 nm, with increased standard deviations. The largest standard deviations in the monthly effective SSA were found at nm. In general, the uncertainties became higher in the UV-B spectral range, which was expected. The monthly Brewer effective SSA averages ranged from 0.81 to 0.86 at nm, while at and 320 nm all values were above 0.90, therefore the aerosols absorbance seems to

147 147 increase with wavelength. However, the estimations at UV-B wavelengths are subject to greater influence due to increased measurement uncertainties and the effect of ozone absorption. A possible future step of this work would be the examination of the possible increased absorption of aerosols during specific days with the use of ancillary data (e.g. aerosol chemical composition, back-trajectory analysis, lidar observations) in the frame of experimental campaigns dedicated in aerosol optical properties in the UV region Conclusions The measurements of AOD, total ozone and UV irradiance from the Brewer#178 instrument at Uccle, Belgium, were used, for the first time, to estimate the aerosol RFE in the UV range at a typical site in Western Europe. The AOD values, retrieved from the Brewer direct sun measurements using the Langley Plot Method, were in very good agreement with the extrapolated ones at the same wavelength from the collocated CIMEL for the period of July 2006-May The aerosol RFE increased for decreasing sza. In the nm spectral region, the highest values were revealed at 30 o (-6.9 ± 0.9 W/m 2 ), while at 60 o the RFE was almost 2.5 times lower (-2.7 ±0.1 W/m 2 ). In the UV-B region ( nm), the RFE value at 60 o ( ±0.005 W/m 2 ) was 5 times lower than the corresponding value at 30 o (-0.35 ±0.04 W/m 2 ). A short decrease is revealed also for the RFE(%) with increasing sza, which is within the measurement uncertainties. Model calculations of global irradiance were also been used along with the Brewer UV measurements to estimate the aerosol effective SSA at five UV wavelengths (306.5, 310, 313.5, and 320 nm). The novelty in this study was that, taking advantage of the AOD measurements at these wavelengths, the validity of the method was examined, for the first time in Brewer spectral measurements, at low UV-A wavelengths and in the UV-B region. The estimated Brewer effective SSA monthly averages at 320 nm are in very close agreement (within ±0.01) with the CIMEL measurements at 440nm. Due to increased measurement uncertainties and the effect of ozone absorption, larger standard deviations were calculated for the monthly effective SSA, at nm. For the rest of wavelengths, the calculated standard deviation was comparable to that at 320 nm.

148 148 References Antón, M., Gil, J.E., Fernández-Gálvez, J., Lyamani, H., Valenzuela, A., Foyo-Moreno, I., Olmo, F.J., Alados-Arboledas, L., Evaluation of the aerosol forcing efficiency in the UV erythemal range at Granada, Spain. Journal of Geophysical Research 116, D doi: /2011jd Antón, M., Sorribas, M., Bennouna, Y., Vilaplana, J.M., Cachorro, V.E., Gröbner, J., Alados-Arboledas, L., Effects of an extreme desert dust event on the spectral ultraviolet irradiance at El Arenosillo (Spain). Journal of Geophysical Research 117, D03205, doi: /2011jd Bais, A., Kazantzidis, A., Kazadzis, S., Balis, D.S., Zerefos, C.S., Meleti, C., Deriving an effective aerosol single scattering albedo from spectral surface UV irradiance measurements. Atmospheric Environment 39(6), Bais, A.F., Tourpali, K., Kazantzidis, A., Akiyoshi, H., Bekki, S., Braesicke, P., Chipperfield, M.P., Dameris, M., Eyring, V., Garny, H., Iachetti, D., Jöckel, P., Kubin, A., Langematz, U., Mancini, E., Michou, M., Morgenstern, O., Nakamura, T., Newman, P.A., Pitari, G., Plummer, D.A., Rozanov, E., Shepherd, T.G., Shibata, K., Tian, W., Yamashita, Y., Projections of UV radiation changes in the 21 st century: impact of ozone recovery and cloud effects. Atmospheric Chemistry and Physics 11, Bass, A.M., Paur, R.J., The ultraviolet cross-sections of ozone: I. The measurements, Atmospheric Ozone. Ed by C.S. Zerefos and A. Ghazi, Proc. Quadrennial Ozone Symposium, Reidel, Dordrecht, Germany, Bäumer, D., Rinke, R., Vogel, B., Weekly periodicities of aerosol optical thickness over Central Europe evidence of an anthropogenic direct aerosol effect. Atmospheric Chemistry and Physics 8, doi: /acp Cachorro, V.E., Berjón. A., Toledano, C., Mogo, S., Prats, N., de Frutos, A.M., Vilaplana, J.M., Sorribas, M., de la Marena, B.A., Gröbner, J., Lalulainen, N., Detailed Aerosol Optical Depth Intercomparison between Brewer and Li-Cor 1800 Spectroradiometers and a Cimel Sun Photometer. Journal of Atmospheric and Oceanic Technology 26,

149 149 Cheymol, A., De Backer, H., Retrieval of the aerosol optical depth in the UV-B at Uccle from Brewer ozone measurements over a long time period Journal of Geophysical Research 108 (D24), doi: /2003JD Cheymol, A., Gonzalez Sotelino, L., Lam, K.S., Kim, J., Fioletov, V., Siani, A.M., De Backer, H., Intercomparison of Aerosol Optical Depth from Brewer Ozone spectrophotometers and CIMEL sunphotometers measurements. Atmospheric Chemistry and Physics 9, Cordero, R.R., Seckmeyer, G., Pissulla, D., Dasilva, L., Labbe, F., Uncertainty evaluation of the spectral UV irradiance evaluated by using the UVSPEC radiative transfer model. Optics Communications 276, Corr, C.A., Krotkov, N., Madronich, S., Slusser, J.R., Holben, B., Gao, W., Flynn, J., Lefer, B., Kreidenweis, S.M., Retrieval of aerosol single scattering albedo at ultraviolet wavelengths at the T1 site during MILAGRO. Atmospheric Chemistry and Physics 9, De Backer, H., Time series of daily erythemal UV doses at Uccle, Belgium. International Journal of Remote Sensing 30, 15-16, De Bock, V., De Backer, H., Mangold, A., Delcloo, A., Aerosol Optical Depth measurements at 340 nm with a Brewer spectrophotometer and comparison with Cimel sunphotometer observations at Uccle, Belgium. Atmospheric Measurement Techniques 3, den Outer, P.N., Slaper, H., Kaurola, J., Lindfors, A., Kazantzidis, A., Bais, A.F., Feister, U., Junk, J., Janouch, M., Josefsson, W., Reconstructing of erythemal radiation levels in Europe for the last 4 decades. Journal of Geophysical Research 115, D Díaz, A.M., García, O.E., Díaz, J.P., Expósito, F.J., Utrillas, M.P., Martínez-Lozano, J.A., Alados- Arboledas, L., Olmo, F.J., Lorente, J., Cachorro, V., Horvath, H., Labajo, A., Sorribas, M., Vilaplana, J.M., Silva, A.M., Elias, T., Pujadas, M., Rodrigues, J.A., González, J.A., Aerosol radiative forcing efficiency in the UV region over southeastern Mediterranean: VELETA2002 campaign. Journal of Geophysical Research 112, D doi: /2006jd

150 150 Di Biagio, C., di Sarra, A., Meloni, D., Monteleone, F., Piacentino, S., Sferlazzo, D., Measurements of Mediterranean aerosol radiative forcing and influence of the single scattering albedo. Journal of Geophysical Research 114, D doi: /2008jd Di Biagio, C., di Sarra, A., Meloni, D., Large atmospheric shortwave radiative forcing by Mediterranean aerosols derived from simultaneous ground-based and spaceborne observations and dependence on the aerosol type and single scattering albedo. Journal of Geophysical Research 115, D doi: /2009jd Garcίa, O.E., Dίaz, A.M., Expósito, F.J., Dίaz, J.P., Gröbner, J., Fioletov, V.E., Cloudless aerosol forcing efficiency in the UV region from AERONET and WOUDC databases. Geophysical Research Letters 33, L doi: /2006gl Georgoulias, A.K., Kourtidis, K.A., On the aerosol weekly cycle spatiotemporal variability over Europe. Atmospheric Chemistry and Physics 11, Gröbner, J., Meleti, C., Aerosol optical depth in the UVB and visible wavelength range from Brewer spectrophotometer direct irradiance measurements: Journal of Geophysical Research 109, D doi: /2003JD Gröbner J., Blumthaler M., Kazadzis S., Bais A., Webb A.R., Schreder J., Seckmeyer G., Rembges D., Quality assurance of spectral solar UV measurements: results from 25UV monitoring sites in Europe, 2002 to Metrologia 43, S66 S71. doi: / /43/2/s14. Hansen, J.E., Sato, M., Ruedy, R., Radiative forcing and climate response. Journal of Geophysical Research 102, Herman, J.R., Celarier, E.A., Earth surface reflectivity climatology at nm from TOMS data. Journal of Geophysical Research 102 (D23), Holben, B.N., Eck, T.F., Slutsker, I., Tanré, D., Buis, J.P., Setzer, A., Vermote, E., Reagan, J.A., Kaufman, Y.J., Nakajima, T., Lavenu, F., Jankowiak, I., Smirnov, A., AERONET A federated instrument network and data archive for aerosol characterization. Remote Sensing of Environment 66 (1), Holben, B.N., Tanré, D., Smirnov, A., Eck, T.F., Slutsker, I., Abuhassan, N., Newcomb, W.W., Schafer, J., Chatenet, B., Lavenue, F., Kaufman, Y.J., Vande Castle, J., Setzer, A., Markham, B., Clark, D., Frouin, R.,

151 151 Halthore, R., Karnieli, A., ONeill, N.T., Pietras, C., Pinker, R.T., Voss, K., Zibordi, G., An emerging ground-based aerosol climatology: Aerosol Optical Depth from AERONET. Journal of Geophysical Research 106, Ialongo, I., Buchard, V., Brogniez, C., Casale, G.R., Siani, A.M., Aerosol Single Scattering Albedo retrieval in the UV range: an application to OMI satellite validation. Atmospheric Chemistry and Physics 10, Intergovernmental Panel on Climate Change (IPCC): Climate Change 2007, Working Group I Report: The physical science basis, Summary for Policymakers, Paris, February Karnieli, A., Derimian, Y., Indoitu, R., Panov, N., Levy, R.C., Remer, L.A., Maenhaut, W., Holben, B.N., Temporal trend in anthropogenic sulfur aerosol transport from central and eastern Europe to Israel. Journal of Geophysical Research 114, D00D19. doi: /2009jd Kazadzis, S., Bais, A., Amiridis, V., Balis, D., Meleti, C., Kouremeti, N., Zerefos, C.S., Rapsomanikis, S., Petrakakis, M., Kelesis, A., Tzoumaka, P., Kelektsoglou, K., Nine years of UV aerosol optical depth measurements at Thessaloniki, Greece. Atmospheric Chemistry and Physics 7, Kazadzis, S., Kouremeti, N., Bais, A., Kazantzidis, A., Meleti, C., Aerosol forcing efficiency in the UVA region from spectral solar irradiance measurements at an urban environment. Annales Geophysicae 27, Kazadzis, S., Gröbner, J., Arola, A., Amiridis, V., The effect of the global UV irradiance measurement accuracy on the single scattering albedo retrieval. Atmospheric Measurement Techniques 3, Kazantzidis, A., Balis, D.S., Bais, A.F., Kazadzis, S., Galani, E., Kosmidis, E., Blumthaler, M., Comparison of model calculations with spectral UV measurements during the SUSPEN campaign: the effects of aerosols. Journal of Atmospheric Sciences 58, Kazantzidis, A., Bais, A.F., Balis, D.S., Kosmidis, E., Zerefos, C.S., Sensitivity of solar UV radiation to ozone and temperature profiles at Thessaloniki (40.5 N, 23 E), Greece. Journal of Atmospheric and Solar-Terrestrial Physics 67,

152 152 Koelemeijer, R.B.A., Homan, C.D., Matthijsen, J., Comparison of spatial and temporal variations of aerosol optical thickness and particulate matter over Europe. Atmospheric Environment 40, Koukouli, M.E., Kazadzis, S., Amiridis, V., Ichoku, C., Balis, D.S., Bais, A.F., Signs of a negative trend in the MODIS aerosol optical depth over the Southern Balkans. Atmospheric Environment 44, Kylling, A., Bais A.F., Blumthaler M., Schreder J., Zerefos C.S., Kosmidis E., Effect of aerosols ol solar IJV irradiances during the Photocheimical Activity and Solar Ultraviolet Radiation campaign. Journal of Geophysical Research 103 (1D20), Lakkala, K., Arola, A., Heikkilä, A., Kaurola, J., Koskela, T., Kyrö, E., Lindfors, A., Meinander, O., Tanskanen, A., Gröbner, J., Hülsen, G., Quality assurance of the Brewer spectral UV measurements in Finland. Atmospheric Chemistry and Physics 8, Lee, K.H., Li, Z., Wong, M.S., Xin, J., Wang, Y., Hao, W.M., Zhao, F., Aerosol single scattering albedo estimated across China from a combination of ground and satellite measurements. Journal of Geophysical Research 112, D22S15. doi: /2007JD Marenco, F., di Sarra, A., Di Luisi, J., Methodology for determining aerosol optical depth from Brewer nm ozone measurements. Applied Optics 41(9), doi: /AO Mayer, B., Kylling, A., Technical note: The libradtran software package for radiative transfer calculations description and examples of use. Atmospheric Chemistry and Physics 5, McKenzie, R.L., Aucamp, P.J., Bais, A.F., Bjorn, L.O., Ilyas, M., Madronich. S., Ozone depletion and climate change: impacts in UV radiation. Photochemical & Photobiological Sciences 10, 2, Meleti, C., Cappellani, F., Measurements of aerosol optical depth at Ispra: Analysis of the correlation with UV-B, UV-A and total solar irradiance. Journal of Geophysical Research 105(D4), Nikitidou, E., Kazantzidis, A., De Bock, V., De Backer, H., The aerosol forcing efficiency in the UV region and the estimation of single scattering albedo at a typical West European site. Atmospheric Environment 69,

153 153 Petters, J.L., Saxena, V.K., Slusser, J.R., Wenny, B.N., Madronich, S., Aerosol single scattering albedo retrieved from measurements of surface UV irradiance and a radiative transfer model. Journal of Geophysical Research 108(D9), doi: /2002jd Sellitto, P., di Sarra, A., Siani, A.M., An improved algorithm for the determination of aerosol optical depth in the ultraviolet spectral range from Brewer spectrophotometer observations. Journal of Optics A: Pure and Applied Optics 8, Shettle, E.P., Models of aerosols, clouds and precipitation for atmospheric propagation studies. Paper presented at Conference on Atmospheric Propagation in the UV, Visible, IR and MM-Region and Related System Aspects, NATO Adv. Group for Aerosp. Res. and Dev., Copenhagen, Streets, D.G., Yan, F., Chin, M., Diehl, T., Mahowald, N., Schultz, M., Wild, M., Wu, Y., Yu, C., Anthropogenic and natural contributions to regional trends in aerosol optical depth, Journal of Geophysical Research 114, D00D18. doi: /2008jd Tourpali, K., Bais, A.F., Kazantzidis, A., Zerefos, C.S., Akiyoshi, H., Austin, J., Brühl, C., Butchart, N., Chipperfield, M.P., Dameris, M., Deushi, M., Eyring, V., Giorgetta, M.A., Kinnison, D.E., Mancini, E., Marsh, D.R., Nagashima, T., Pitari, G., Plummer, D.A., Rozanov, E., Shibata, K., Tian, W., Clear sky UV simulations for the 21 st century based on ozone and temperature projections from Chemistry-Climate Models. Atmospheric Chemistry and Physics 9, Van Dijk, A., Slaper, H., den Outer, P.N., Morgenstern, O., Braesicke, P., Pyle, J.A., Garny, H., Stenke, A., Dameris, M., Kazantzidis, A., Tourpali, K., Bais. A.F., Skin cancer risks avoided by the Montreal protocol-worldwide modeling integrating couple climate-chemistry models with a risk model for UV. Photochemical & Photobiological Sciences. doi: /j x. Vasaras, A., Bais, A.F., Feister, U., Zerefos, C.S., Comparison of two methods for cloud flagging of spectral UV measurements. Atmospheric Research 57(1), Wilawan, K., Rimmer, J.S., Smedley, A.R.D., Ying T.Y., Webb A.R., Aerosol Optical Depth and the Global Brewer Network: A Study Using U.K.- and Malaysia-Based Brewer Spectrophotometers. Journal of Atmospheric and Oceanic Technology 29,

154 154 Xia, X., Eck, T.F., Holben, B.N., Phillippe, G., Chen, H., Analysis of the weekly cycle of aerosol optical depth using AERONET and MODIS data. Journal of Geophysical Research 113, D doi: /2007jd Yoon, J. von Hoyningen-Huene, W., Vountas M., Burrows J.P., Analysis of linear long-term trend of aerosol optical thickness derived from SeaWiFS using BAER over Europe and South China. Atmospheric Chemistry and Physics 11, Zerefos, C.S., Eleftheratos, K., Meleti, C., Kazadzis, S., Romanou, A., Ichokou, C., Tselioudis G., Bais, A., Solar dimming and brightening over Thessaloniki, Greece, and Beijing, China. Tellus 61B, Zhang J., Reid J.S., A decadal regional and global trend analysis of the aerosol optical depth using a data-assimilation grade over-water MODIS and Level 2 MISR aerosol products. Atmospheric Chemistry and Physics 10, Zhao, T.X.-P., Laszlo, I., Guo, W., Heidinger, A., Cao, C., Jelenak, A., Tarpley, D., Sullivan, J., Study of long-term trend in aerosol optical thickness observed from operational AVHRR satellite instrument. Journal of Geophysical Research 113, D doi: /2007jd

155 Aerosol effects on Direct Normal Irradiance in Europe Abstract The effect of spatial and temporal variability of aerosol optical depth (AOD) on direct normal irradiance (DNI) under cloud-free skies is studied, with the synergetic use of satellite and ground-based data as well as calculations from a radiative transfer model. The area of interest is Europe and data from May to September during 13 years ( ) are analyzed. The aerosol effect on DNI is high in areas influenced by desert dust intrusions and intense anthropogenic activities, such as the Mediterranean basin and the Po Valley in Italy. In May, the attenuation of DNI from aerosols, over these areas, can reach values up to 35% and 45% respectively, which corresponds to 4 and 6 kwh/m 2 per day. In most areas, even for periods with lower values of AOD, the attenuation on DNI is found to be around 20%, which corresponds to about 2 to 3 kwh/m 2 less received DNI per day. However, the DNI has increased during the recent years, due to the decreasing tendency of AOD over most areas of Europe. The increase is around 6 to 12%, which corresponds to an amount of 0.5 to 1.25 more kwh/m 2 received per day. The percentage differences of daily DNI from the corresponding monthly climatological value reveals that day-to-day differences (due to AOD changes) from the monthly mean, up to ±20%, can occur. The significance of the aerosol changes in Europe reveals the necessity for near real-time measurements or forecasts of AOD when reliable estimations of DNI are required Introduction Aerosols are one of the most important constituents in the atmosphere that affect the incoming solar radiation, either directly through absorbing and scattering processes or indirectly by changing the optical properties and lifetime of clouds. Even though aerosols have been in the center of scientific research for 3 Based on publication: Nikitidou et al., 2013, submitted in Renewable Energy

156 years, their high spatiotemporal variability and complex atmospheric interactions induce challenges in the determination of their radiative forcing, which still holds high uncertainties (IPCC, 2007). 156 Many studies focus on the spatiotemporal distribution of aerosol properties in Europe. Chubarova (2009) studied the seasonal distribution, using data from the MODIS instrument and the AERONET network. The highest aerosol loads were observed in south and southeastern parts of Europe during the warmer months of the year. Over these areas, permanent high values of the Ångström α coefficient were found indicating the presence of anthropogenic aerosols. Marmer and Langmann (2007) studied the interannual variability of aerosol distribution in Europe, using a regional atmosphere-chemistry model. They highlighted the role of meteorological conditions which can produce up to 100% variability in the monthly mean aerosol load. The weekly variability of aerosols in Europe presents the lowest values during the weekend due to the decrease in transport and other anthropogenic activities (Baumer et al., 2008, Xia et al., 2008, Georgoulias and Kourtidis, 2011). As a result of legislations regarding atmospheric pollution, negative trends in aerosol amounts have been observed during the last years in many areas of Europe, the so-called brightening effect (Papadimas et al., 2008, Karnieli et al., 2009, Streets et al., 2009, Koukouli et al., 2010, Yoon et al., 2011). In order to complement the ground measurements and account for their limited spatial coverage, satellites have been employed for Earth s and atmosphere s observation, which can provide global coverage and produce long-time datasets. The MODerate resolution Imaging Spectroradiometer (MODIS), onboard the NASA Terra satellite, has been in operation since February The sun-synchronous satellite provides global coverage every 1-2 days and acquires data in 36 spectral bands ( Remer et al., 2005). The Collection 005 (C005) is the dataset of MODIS products provided by the version V5.2 of the algorithm, which replaced previously used versions (Levy et al., 2007a, 2007b). The prelaunch expected error in AOD retrievals over land is ± ( AOD), while AOD is estimated to within this expected error in more than 60% and 72% of the cases over ocean and over land respectively (Remer et al., 2008). The MODIS data have been extensively validated against ground-based measurements (Chu et al., 2002, Santese et al., 2007, Levy et al., 2010, De Meij et al., 2011) and used to calculate, amongst others, the

157 157 aerosol trends and events (Papadimas et al., 2008, Gkikas et al., 2009, Zhang and Reid, 2010), as well as the surface shortwave (Bisht et al., 2005, Tang et al., 2006, Huang et al., 2011) ultraviolet and visible solar irradiance (Nikitidou et al., 2012). The accurate knowledge of the amount of solar irradiance reaching a surface and its temporal variability is essential for the efficient performance of solar power applications. The most important parameter for these systems is the Direct Normal Irradiance (DNI), which is the radiation received by a surface perpendicular to the direction of the sun. Although systems measuring the Global Horizontal Irradiance (GHI) are abundant, DNI measurements are not so common and these data are usually estimated indirectly with the use of radiative transfer or decomposition models (Battles et al., 2000, Lopez et al., 2000, Janjai et al., 2011). Under cloud-free skies, aerosols become the dominant factor that affect the amount of solar irradiance reaching the ground. It has been shown that the variability in DNI due to aerosols is more important than the one induced in GHI (Gueymard et al., 2012), while the uncertainty in its calculation is dominated by uncertainties in aerosol optical properties (Halthore et al., 1997). Suri et al., (2009) studied the DNI in Europe as this is provided by 5 different datasets. The annual sum is found to reach the highest values in areas of the Mediterranean, Southern and Central Spain, Portugal, Sicely, Sardinia and Provence. Gueymard (2012) used AOD measurements at 180 stations of the AErosol RObotic NETwork (AERONET) over the world to study the variability of AOD and its effect on DNI and GHI. He concluded that some areas experience too high variability that makes resource assessments potentially too optimistic for bankability if based only on limited data series. In this study, the effect of AOD on DNI under cloud-free skies is studied, with the synergetic use of satellite data from the MODIS instrument, complementary data from the AERONET network and calculations from a radiative transfer model. The area of interest is Europe and data from a 13-year time period ( ) are analyzed. The DNI temporal and spatial variability due to aerosols is examined, as well as the increase induced by the observed decreasing tendency of AOD over Europe. The uncertainties of DNI estimations induced by the use of aerosol climatological values are investigated.

158 Data and Methodology The AOD at 550 nm is taken from the MODIS Terra daily Level-3 data (Collection 5.1) which have a spatial resolution of 1 x1 (100x100 km). Terra was launched on December 1999 and started providing MODIS data on March The MODIS AOD at 550 nm, from the Terra satellite is used in this study from the beginning of operation (March 1 st, 2000) until the end of The AOD wavelength dependence, described by the α Ångström exponent, is provided by MODIS but is not used in this study since it still presents considerable errors when compared with ground-based measurements (Levy et al., 2007b, 2010). For the Ångström α exponent, the level 2.0 climatological data from AERONET were used. The level 2.0 data are pre- and post-field calibrated, automatically cloud screened and manually checked. The monthly climatological values of Ångstrom α exponent ( nm) were chosen for 37 stations in Europe, which present data availability higher than 3 years during the period These data were then interpolated for the region of our study in order to obtain the spatial resolution of MODIS (1 x1 ). The interpolation was performed using the method of Ordinary Kriging, combined with a linear semi-variogram model. The AERONET sites used in this study are shown in Figure 5.1. The area examined (latitude: 29.5 N 59.5 N, longitude: 9.5 W 38.5 E) covers Europe, with the exception of the northern latitudes of Scandinavia, the north coasts of Africa and the Mediterranean coasts of Middle East.

159 159 Figure 5.1: AERONET stations used for the retrieval of the gridded monthly averages of Ångström α exponent. In order to estimate the DNI on the surface, the radiative transfer model SBDART is used (Ricchiazzi et al., 1998), which is included in the LibRadtran package (Mayer and Kylling, 2005, Typical for the mid-latitudes vertical profiles for the basic atmospheric gases, pressure and temperature are used (Anderson et al., 1986). The surface albedo is set at 0.2 for the shortwave (SW) range ( nm) (Csiszar and Gutman, 1999). The aerosol vertical profile is described by Shettle (1989). The AOD is described by the Ångström α and β exponents, the latter of which is calculated by the equation: β = τ * λ α, (5.1) where τ is the MODIS AOD at λ = 550 nm and α the AERONET Ångström exponent. The solver of the radiative transfer equation is the discrete ordinate algorithm with 6 streams considered and the sphericity of the Earth is taken into account. The AOD provided by MODIS Terra is considered to be constant during the day. For every satellite pixel and day, the cloud-free DNI is calculated every 15 minutes, from sunrise to sunset, as a function of the

160 160 solar zenith angle, the α and β Ångström exponents. A correction for the Earth Sun distance is subsequently applied. The daily received energy from DNI is then calculated by integrating on the diurnal values. The DNI is also calculated for aerosol-free conditions in order to derive the absolute and percentage (%) differences in the received DNI between the measured and the aerosol-free conditions: Diff DNI (5.2) DNI clear DNI DNIclear Diff (%) ( )*100, (5.3) DNI clear where DNI clear is the calculated DNI under aerosol-free conditions. The latitudes lower than 34 are not depicted in the maps of the following figures, as the satellite provides few data during the investigated period due to the high surface albedo of the desert areas. During the 13-year period examined, only months which had data availability higher than 40% (more than 12 days available data during each month) were considered. As a result, only data during the period May September of each year are analyzed, since the data availability during the rest of the months is not adequate due to the dominance of thick cloudiness. All results refer to the daily energy, calculated by integrating on the diurnal DNI values Results The Ångström α exponent seasonal variability The Ångström α exponent monthly climatological values, taken from the selected AERONET stations in Europe, were interpolated to match the MODIS spatial resolution (1 x1 ). The monthly Ångström

161 α exponent maps for Europe during the May-September period are presented in Figure Figure 5.2: Monthly average maps of the Ångström α exponent over Europe. The lowest values (around 0.8) are presented during May in Southern Europe, Mediterranean and the north coasts of Africa and are evident of the frequent desert dust intrusions during this period. The influence

162 162 of desert dust is very important in July and August too, for Southern Spain and the west coasts of Africa but is limited in Eastern Mediterranean. The highest values of Ångström α exponent ( ) are observed during summer in Central-Eastern Europe and the Balkans, indicating the presence of fine mode particles, which persist during this period due to the reduction of precipitation. In September, these values decrease and moderate values are observed in continental Europe, while maritime areas are described by values around DNI changes due to the AOD variability The aerosol effect on DNI The monthly mean values of AOD, at 550 nm, are presented in Figure 5.3. Only the months May Septermber are presented, as they fulfil the data availability requirements, mentioned in the previous section. The highest AOD values are observed in the Po Valley, during the months of May and June, as well as in the South Western Mediterranean in July. Central Turkey also exhibits high aerosol loads during the investigated period. In September, the aerosol amounts decrease in all areas, though Po Valley and Central Turkey still have considerably high values, around The absolute and percentage differences of DNI under cloud-free skies have been calculated for every day and every satellite pixel during the examined period. The monthly average values of the percentage difference over the 13-year period are presented in Figure 5.4.

163 Figure 5.3: Monthly mean values of AOD at 550 nm. Only the priod May-September is presented, which fulfils the data availability requirements. 163

164 164 Figure 5.4: Monthly mean values of the aerosol effect (%) on DNI. Only the period May-September is presented, which fulfils the data availability requirements. In Mediterranean, the highest percentage differences are observed in May reaching values down to -35%. This corresponds to an attenuation of about 4 kwh/m 2 per day. The highest reductions during this month are evident around Po Valley, with differences down to -45% (-6 kwh/m 2 per day). Lower values of -25 to -

165 165 15% (-3 to -2 kwh/m 2 per day) are revealed in Spain, Southern France, Eastern and Northern Europe, while Central Europe and the southern parts of Ireland and Britain present an attenuation of around -30 to -25% (-4 to -3.5 kwh/m 2 per day). The data availability at the northern parts of Ireland and Britain was lower than 40% so the results for these areas are considered to be of no significance for this study. In June, the percentage difference is slightly lower over the Mediterranean basin and Northern Europe but moderately increases in areas of central Europe. In Northern France, the Low Countries, most parts of Germany and Poland, the attenuation in DNI is around 30-35%, while it remains high around the Po Valley and the far western part of the Mediterranean basin. The situation in continental Europe is similar in July, with the exception of a slight increase of the aerosol effect in North-East regions. In the Mediterranean, the effect of aerosols on DNI is decreased over Portugal and Po Valley (around -10 and -30% respectively) but becomes stronger over the west coasts of Africa, where percentage differences up to -40% (-5 kwh/m 2 per day) are revealed. In August, the aerosol effect on DNI is around -25 to -30% (-3 to -3.5 kwh/m 2 per day) over most parts of Mediterranean, Central and Eastern Europe and slightly lower in France and Spain, where differences around -2 kwh/m 2 per day are calculated. The effect is stronger in South-Western Mediterranean and in the interior areas of Turkey. In September, the percentage differences decrease in Central and East Europe, while there is an increase in the North Sea (-25 to -35%). The importance of the aerosol effect on DNI is clear, as even for months with low aerosol loads, the attenuation of DNI is around -20%, which corresponds to about -2 to -3 kwh/m 2 per day. This effect becomes even more significant at areas burdened with aerosols throughout most time of the year, such as the industrial region of Northern Italy Annual Differences of DNI For each year, the average values of modeled DNI have been calculated over the period May September. Figure 5.5 presents the percentage differences, for every year, between these mean values and the average from the period May September of all years.

166 Figure 5.5: Differences (%) between the mean DNI for each year ( ) and the average for the 13-year period. Only data for May-September are considered. 166

167 167 There is an increasing tendency in the calculated differences. During the first years of the period examined, the differences in DNI are negative; they reach low values, around zero, in the middle of the period and then increase during the last years, indicating a positive trend in the received DNI due to the reduction of AOD ( brightening effect). This brightening effect is also affected by changes in aerosol optical properties, besides changes in their number concentration. During the years 2000 to 2003, the DNI received during the May September period is lower than the average derived from the entire 13-year period, for most regions of Europe. Central Europe receives around 4 to 10% lower DNI during 2000 and 2001, which corresponds to -1 to kwh/m 2 per day. However, positive differences (2-4%) are revealed in some areas of Northern Europe. In 2002, Eastern Europe and the Balkans present the highest negatives differences, as 8 to 18% (0.75 to 1.75 kwh/m 2 per day ) less DNI was received compared to the average mean probably due to excessive wildfires in this area (Ovadnevaitė et al., 2006, Hänniken et al., 2009). In 2003, low negative values ( -10 to -4%, -1 to kwh/m 2 per day) are observed in West-Central Europe and Western Mediterranean due to exceptionally hot and dry conditions and large-scale intense forest fires (Pace et al., 2005, Garcia-Herrera et al., 2010, Péré et al., 2011). In 2004, small increases start to appear, which reach +2 to +8%, in 2005, in latitudes between 50 and 55 N. From 2006 to 2008, the differences are small, around ±4%, which corresponds to an aerosol effect on DNI less than 0.5 kwh/m 2 per day. However, increasing values of DNI are revealed during the last years ( ) of the study period. In 2012, 6 to 12% more DNI is received on most areas of Central and Eastern Europe, the Balkans and the Italian peninsula. These positive differences correspond to an increase of 0.5 to 1.25 kwh/m 2 in received DNI per day. These differences are further studied in the next section by averaging them over specific regions of Europe DNI and AOD differences per area The differences in DNI discussed above, which were provided by pixel, are now averaged over each of 5 different areas in Europe. These areas are Central-Western Mediterranean (latitude: 34.5 N 44.5 N, longitude: 9.5 W 18.5 E), Eastern Mediterranean-Black Sea (latitude: 32.5 N 46.5 N, longitude: 18.5 E

168 E), Central-Western Europe (latitude: 44.5 N 54.5 N, longitude: 4.5 W 14.5 E), Eastern Europe (latitude: 44.5 N 55.5 N, longitude: 14.5 E 38.5 E) and Northern Europe (latitude: 54.5 N 59.5 N, longitude: 5.5 E 29.5 E) (Figure 5.6). Figure 5.6: Europe divided into 5 areas of interest: Central-Western Mediterranean (34.5 N 44.5 N, 9.5 W 18.5 E), Eastern Mediterranean-Black Sea (32.5 N 46.5 N, 18.5 E 36.5 E), Central-Western Europe (44.5 N 54.5 N, 4.5 W 14.5 E), Eastern Europe (44.5 N 55.5 N, 14.5 E 38.5 E) and Northern Europe (54.5 N 59.5 N, 5.5 E 29.5 E). The percentage differences of the annual mean AOD and DNI values, for the May-September period, from the climatological values over the whole period ( ) are calculated for each area (Figure 5.7). Figure 5.8 presents the same differences, provided in the corresponding parameter units. The AOD annual mean values (Figure 5.7a) are higher than the 13-year average, until 2003, after which a decrease starts, which becomes more evident during the end of the period. During , the AOD in Central-Western

169 169 Europe and Central-Western Mediterranean is lower by 15 17% (~0.04) than the 13-year average, while in Eastern Europe and Eastern Mediterranean a similar decrease is revealed only in 2011 and With the exception of year 2002, where strong positive differences are observed, the area of Northern Europe has the lowest variability during the examined period. The corresponding percentage differences in DNI are presented in Figure 5.7b. Over all regions, the decrease during the beginning of the examined period and the increase at the end is evident. With the exception of Northern Europe, all areas received around 3% (0.3 kwh/m 2 per day) less DNI during 2000 and In 2002, the highest decrease is revealed in Eastern Europe (~10%, 0.9 kwh/m 2 per day). The differences are moderate for all regions in the following years but increased DNI values are revealed after During the end of the examined period, an increase in the received DNI of 4 to 6% ( kwh/m 2 per day) can be observed in all regions, except Northern Europe. difference in AOD (%) a Central-Western Europe Central-Western Mediterranean Eastern Europe Eastern Mediterranean-Black Sea Northern Europe b year 4 difference in DNI (%) year Figure 5.7: Difference (%) between the mean AOD (5.7a) and DNI (5.7b) value for each year ( ) and the average for the 13- year period for the 5 areas of interest.

170 170 Figure 5.8 : Difference (in corresponding units) between the mean AOD (5.8a) and DNI (5.8b) value for each year ( ) and the average for the 13-year period for the 5 areas of interest. Figure 5.9 presents the percentage differences of daily DNI from the corresponding monthly climatological value, calculated from the daily DNI values during the entire time period. The positive trend in DNI differences is evident in all areas, when looking at the daily differences as well, except Northern Europe. Additionally, in all areas, day-to-day percentage differences of DNI, due to changes of AOD, up to ±20% from the monthly climatological value are revealed. The high negative differences which were observed in Eastern Europe, in 2002, can now be seen that occur during the period of mid August mid September. During this time, the differences can reach values higher than -40%. It is during the same period of 2002 that high differences in Northern Europe appear as well.

171 171 Figure 5.9: Difference (%) between the daily DNI and the corresponding monthly mean, for each of the 5 areas of interest DNI derived from MODIS AOD climatology In this section, the monthly climatological MODIS AOD values, derived from the 13-year period, have been used as input data in the SBDART model, in order to estimate the DNI values. The goal is to examine the differences between these DNI values and the monthly ones, calculated from the available daily data during the examined period. The analysis is performed for the months May to September and the results are averaged over each of the 5 areas described before. Figure 5.10 shows the differences between the DNI derived from the monthly AOD climatology and the corresponding monthly mean DNI calculated from all the daily AOD data during the period in study.

172 172 Figure 5.10: Difference (%) between the DNI derived from the monthly AOD climatology and the corresponding monthly mean DNI calculated from the available daily AOD data during the period and over each of the 5 areas of interest. The derived differences are negative for all months and all regions, showing that when the AOD monthly climatology is used to estimate the DNI, there is an underestimation of a few percent, as opposed to the DNI monthly values calculated from the daily MODIS data. The lowest differences are observed in Northern Europe in June and July, around 0.16 kwh/m 2 per day, which corresponds to an underestimation of 1.5%. Northern Europe has higher differences for the rest of the months and the maximum difference for the area is found in May (-0.4 kwh/m 2 per day). In Central-Western Mediterranean the underestimation is around 2.5 to 3%, which corresponds to kwh/m 2 per day, except in May when it reaches 0.33 kwh/m 2 per day. In May, the highest, for each region, differences, are found, close to and larger than -0.3 kwh/m 2 per day, except for Eastern Europe which has a slightly higher underestimation in August (0.37 kwh/m 2 per day). These differences are in overall smaller than half a kwh/m 2 per day, but still have to be taken into account if someone chooses to use the satellite climatology for modeling the DNI received by a solar power farm.