How To Understand Cloud Properties From Satellite Imagery



Similar documents
Evaluation of the Effect of Upper-Level Cirrus Clouds on Satellite Retrievals of Low-Level Cloud Droplet Effective Radius

Remote Sensing of Contrails and Aircraft Altered Cirrus Clouds

Advances in Cloud Imager Remote Sensing

Cloud detection and clearing for the MOPITT instrument

Cloud Thickness Estimation from GOES-8 Satellite Data Over the ARM-SGP Site

Clear Sky Radiance (CSR) Product from MTSAT-1R. UESAWA Daisaku* Abstract

GOES-R Advanced Baseline Imager (ABI) Algorithm Theoretical Basis Document For Low Cloud and Fog

Overview of the IR channels and their applications

GOES-R AWG Cloud Team: ABI Cloud Height

Denis Botambekov 1, Andrew Heidinger 2, Andi Walther 1, and Nick Bearson 1

Studying cloud properties from space using sounder data: A preparatory study for INSAT-3D

SAFNWC/MSG Cloud type/height. Application for fog/low cloud situations

Clouds and the Energy Cycle

Volcanic Ash Monitoring: Product Guide

The Effect of Droplet Size Distribution on the Determination of Cloud Droplet Effective Radius

USING THE GOES 3.9 µm SHORTWAVE INFRARED CHANNEL TO TRACK LOW-LEVEL CLOUD-DRIFT WINDS ABSTRACT

REMOTE SENSING OF CLOUD-AEROSOL RADIATIVE EFFECTS FROM SATELLITE DATA: A CASE STUDY OVER THE SOUTH OF PORTUGAL

Night Microphysics RGB Nephanalysis in night time

Comparison between current and future environmental satellite imagers on cloud classification using MODIS

How to Use the NOAA Enterprise Cloud Mask (ECM) Andrew Heidinger, Tom Kopp, Denis Botambekov and William Straka JPSS Cloud Team August 29, 2015

The APOLLO cloud product statistics Web service The APOLLO cloud product statistics Web service

CERES Edition 2 & Edition 3 Cloud Cover, Cloud Altitude and Temperature

The Next Generation Flux Analysis: Adding Clear-Sky LW and LW Cloud Effects, Cloud Optical Depths, and Improved Sky Cover Estimates

Satellite remote sensing using AVHRR, ATSR, MODIS, METEOSAT, MSG

NOAA NESDIS CENTER for SATELLITE APPLICATIONS and RESEARCH ALGORITHM THEORETICAL BASIS DOCUMENT. ABI Cloud Mask

Improved diagnosis of low-level cloud from MSG SEVIRI data for assimilation into Met Office limited area models

IMPACT OF DRIZZLE AND 3D CLOUD STRUCTURE ON REMOTE SENSING OF CLOUD EFFECTIVE RADIUS

A SURVEY OF CLOUD COVER OVER MĂGURELE, ROMANIA, USING CEILOMETER AND SATELLITE DATA

The APOLLO cloud product statistics Web service

Cloud Masking and Cloud Products

Outline of RGB Composite Imagery

Measurement of the effect of biomass burning aerosol on inhibition of cloud formation over the Amazon

Towards assimilating IASI satellite observations over cold surfaces - the cloud detection aspect

Remote Sensing of Clouds from Polarization

Comparison of NOAA's Operational AVHRR Derived Cloud Amount to other Satellite Derived Cloud Climatologies.

NOAA flight planning support and satellite aerosol and cloud retrieval validation during TORERO

Satellite Remote Sensing of Volcanic Ash

DISCRIMINATING CLEAR-SKY FROM CLOUD WITH MODIS ALGORITHM THEORETICAL BASIS DOCUMENT (MOD35) MODIS Cloud Mask Team

Assessing Cloud Spatial and Vertical Distribution with Infrared Cloud Analyzer

The study of cloud and aerosol properties during CalNex using newly developed spectral methods

Validating MOPITT Cloud Detection Techniques with MAS Images

Radiative effects of clouds, ice sheet and sea ice in the Antarctic

CALIPSO, CloudSat, CERES, and MODIS Merged Data Product

Cloud Climatology for New Zealand and Implications for Radiation Fields

A review of cloud top height and optical depth histograms from MISR, ISCCP, and MODIS

Best practices for RGB compositing of multi-spectral imagery

SYNERGISTIC USE OF IMAGER WINDOW OBSERVATIONS FOR CLOUD- CLEARING OF SOUNDER OBSERVATION FOR INSAT-3D

TOPIC: CLOUD CLASSIFICATION

ECMWF Aerosol and Cloud Detection Software. User Guide. version /01/2015. Reima Eresmaa ECMWF

3.4 Cryosphere-related Algorithms

Active Fire Monitoring: Product Guide

Developing Continuous SCM/CRM Forcing Using NWP Products Constrained by ARM Observations

McIDAS-V - A powerful data analysis and visualization tool for multi and hyperspectral environmental satellite data

Validation of SEVIRI cloud-top height retrievals from A-Train data

163 ANALYSIS OF THE URBAN HEAT ISLAND EFFECT COMPARISON OF GROUND-BASED AND REMOTELY SENSED TEMPERATURE OBSERVATIONS

DISCRIMINATING CLEAR-SKY FROM CLOUD WITH MODIS ALGORITHM THEORETICAL BASIS DOCUMENT (MOD35) MODIS Cloud Mask Team

Total radiative heating/cooling rates.

ABI Algorithm Theoretical Basis Document For Daytime Cloud Optical and Microphysical Properties (DCOMP)

SATELLITE OBSERVATION OF THE DAILY VARIATION OF THIN CIRRUS

Hyperspectral Satellite Imaging Planning a Mission

Chapter 6: Cloud Development and Forms

The Surface Energy Budget

CLOUD TOP PROPERTIES AND CLOUD PHASE ALGORITHM THEORETICAL BASIS DOCUMENT

Data processing (3) Cloud and Aerosol Imager (CAI)

Solar Irradiance Forecasting Using Multi-layer Cloud Tracking and Numerical Weather Prediction

RESULTS FROM A SIMPLE INFRARED CLOUD DETECTOR

T.A. Tarasova, and C.A.Nobre

What the Heck are Low-Cloud Feedbacks? Takanobu Yamaguchi Rachel R. McCrary Anna B. Harper

MSG-SEVIRI cloud physical properties for model evaluations

Joint Polar Satellite System (JPSS)

The impact of window size on AMV

DUST DETECTION ALGORITHM USING MODIS DATA AND HYDRA SOFTWARE

Assimilation of cloudy infrared satellite observations: The Met Office perspective

Cloud Oxygen Pressure Algorithm for POLDER-2

CHAPTER 2 Energy and Earth

Retrieval of cloud spherical albedo from top-of-atmosphere reflectance measurements performed at a single observation angle

AATSR Technical Note. Improvements to the AATSR IPF relating to Land Surface Temperature Retrieval and Cloud Clearing over Land

Synoptic assessment of AMV errors

Transcription:

P1.70 NIGHTTIME RETRIEVAL OF CLOUD MICROPHYSICAL PROPERTIES FOR GOES-R Patrick W. Heck * Cooperative Institute for Meteorological Satellite Studies, University of Wisconsin-Madison Madison, Wisconsin P. Minnis Science Directorate, Climate Science Branch NASA Langley Research Center, Hampton, Virginia C. R. Yost Science Systems and Applications, Inc, Hampton, Virginia F.-L. Chang National Institute of Aerospace. Hampton, Virginia A. K. Heidinger NOAA/NESDIS Office of Research and Applications, Camp Springs, Maryland 1. INTRODUCTION Currently, multi-spectral algorithms are being used at NASA Langley Research Center to retrieve microphysical cloud properties from satellite imagery in near-real time over a variety of domains. For nighttime imagery, the Solar infrared-infrared-split window Technique (SIST) is applied to data from the Geostationary Operational Environmental Satellite (GOES) over the Continental US, the Spinning Enhanced Visible InfraRed Imager (SEVIRI) over Europe, the Multifunctional Transport Satellite (MTSAT) over the tropical western Pacific and to Moderate Resolution Imaging Spectroradiometer (MODIS) imagery over the globe. As part of GOES-R Cloud Algorithm Working Group (AWG) efforts at CIMSS, NASA Langleys SIST is being integrated into the GOES-R application framework by removing it from its native environment and revising it to function in the Geostationary Cloud Algorithm Testbed (GEOCAT). This paper presents first results from new SIST retrievals conducted on SEVIRI imagery that is being used as a proxy for GOES-R Advanced Baseline Imager (ABI) data. SIST-derived cloud properties calculated within the GOES-R cloud * Corresponding author address: Patrick W. Heck, CIMSS, University of Wisconsin-Madison, Madison, WI 53706; e-mail: pwheck@ssec.wisc.edu application teams developmental framework are presented. Alterations to SIST, primarily related to using a new option that allows cloud temperature, T cld, and cloud thermodynamic phase to be determined prior to the invocation of SIST, are discussed. 2. DATA Prior to the launch of GOES-R and ABI, the Cloud Algorithm Working Group is utilizing SEVIRI imagery for algorithm development purposes. This study uses results from SEVIRI imagery from August 2006, the Working Groups chosen benchmarking month. The 3.9-, 10.8- and 12- micron channels are used in this preliminary version of the nighttime algorithm. Because the algorithm development is being done within the GEOCAT framework, various ancillary data sets already integrated into that framework are utilized by SIST, as well as results from the various algorithms. Surface characteristics such geotype and wavelength-dependent emissivities from the SeeBor database (Borbas et al., 2007) are obtained via GEOCAT and, notably, the cloud temperature and phase are obtained from baseline algorithms used by the working group. Numerical Weather Predication (NWP) forecast data sets obtained from National Center for Environmental Prediction (NCEP) Global Forecast System (GFS) model data are used to convert cloud temperatures to cloud heights and vice versa, as well as for input to the Radiative Transfer Model (RTM) used within GEOCAT and

for simulating cloud radiances within SIST. The fast RTM currently utilized by the GOES-R Cloud Algorithm Working Group is the PLOD/PFAAST (Hannon et al., 1996). 3. ALGORITHM Because both SIST and its daytime counterpart, the Visible-Infrared-Solar infrared- Split window Technique (VISST) from Minnis et al., 1995, are routinely applied to SEVIRI imagery (see http://www-pm.larc.nasa.gov), transitioning the existing SIST to work in GEOCAT involved extensive alterations to SIST. Several modifications were made in order to exploit SISTs skill in determining cloud optical depth, effective particle size and water path. 3.1. Current Version The current SIST determines the cloud physical parameters that produce the minimum difference between observed and modeled brightness temperature differences for each satellite pixel. Cloud optical depth, effective particle size, water content and temperature of both water and ice clouds are obtained, as well as phase discrimination (Minnis et al., 1995). Observed Brightness Temperature Differences (BTDs) between 3.9 and 10.8, as well as 10.8 and 12 microns, are compared to modeled BTDs and cloud physical parameters are inverted. Those parameters that produce the minimum difference between observed and modeled BTDs for each pixel are assumed to describe the cloud. SIST discriminates between optically thin and thick clouds, but optical depth is reliably derived only for optically thin clouds. In order to derive nighttime cloud properties for each cloudy pixel, the SIST calculates T cld, cloud phase, optical depth and effective cloud particle size with a minimum error, iterative regression method that matches the observations to parameterized emittance calculations (Minnis et al. 1998, Minnis et al. 1995). A 75-term polynomial defines each emittance parameterization which is a function of optical depth, τ, effective ice-crystal diameter D e or water-droplet radius r e, and the temperature difference between the cloud and the surface. BTDs are calculated from the observed satellite radiances such that BTD 34 = T 3.9 µm - T 10.8 µm and BTD 45 = T 10.8 µm - T 12 µm. Atmospheric absorption is accounted for using the correlated k-distribution technique of Kratz (1995). These quantities are computed assuming that all cloudy pixels are overcast and have a temperature given by an initial estimate of T cld = T 10.8 µm over a background with the a priori clearsky temperature for each channel. The phase and a nominal particle size are also assumed for the initial computation. The calculations of BTD 34 and BTD 45 are repeated for both liquid and ice clouds by varying D e or r e, τ and T cld until the difference between the calculated and observed BTDs is minimized. Specifically, we minimize [(BTD 34 - BTD 34 ) 2 + (BTD 45 - BTD 45 ) 2 ]. The selected phase is based on which of the two minimum phase errors is smaller and physically consistent. Cloud height is calculated from the resultant T cld using the appropriate atmospheric profiles. Note that because the radiances observed in a cloudy pixel are assumed to be from a single cloud layer, multi-level scenes or partially cloud-filled pixels can result in less reliable retrievals. 3.2 GOES-R Version For this version of SIST, rather than employing the usual iterative scheme where T cld is allowed to vary along with τ and particle size, T cld is assumed to be previously determined within the GOES-R algorithm. Additionally, the need to simulate BTD 34 and BTD 45 for both ice and water models is eliminated due to the fact that phase cam be determined prior to the invocation of SIST. The elimination of iterating with respect to T cld and calculating solutions for both phases results in a streamlined and faster version of SIST. Within the GOES-R framework, the RTM calculations for each satellite observation have already been made with PLOD/PFAAST and retained, so SIST utilizes those results rather than invoking the correlated k-distribution technique providing consistency with other GOES-R retrievals and additional computational efficiency. 4. CASE STUDIES The first application of the GOES-R SIST was to SEVIRI imagery covering an area off the west-

central coast of Africa. The 10.8-micron imagery (SEVIRI channel 14) is shown in Figure 1. The areas over the ocean are covered by a mixture of Water Supercool Mix Ice Clear Fig. 3. GOES-R AWG derived cloud phase for imagery in Figure 1 that is input to SIST. Fig. 1. SEVIRI infrared imagery (10.8 micron) off the coast of Gabon for 2 August 2006, 0200 UTC. warm clear areas, stratus clouds, both solid and broken, with temperatures around 280º (K) and smaller areas of much colder cirrus clouds over land and sea. In Figures 2 and 3, the input products provided by the GOES-R Cloud AWG and needed for SIST, the derived T cld and phase, are presented. Fig. 2. GOES-R AWG derived cloud temperature for imagery in Figure 1 that is input to SIST. As expected, the stratus cloud temperatures in Fig. 2 are close to the emitted temperatures of Fig. 1, i.e., near 280º(K), but with the impact of partially-filled pixels on the cloud edges apparent. The cirrus cloud temperatures range from about 218º(K) in single layer areas to 250º(K) in multilayer clouds where larger radiances from the warmer stratus below are affecting the retrieved T cld. The accompanying phases in Fig. 3 appear to describe the scene appropriately with water clouds (black) and ice clouds (green) well-chosen. The results from SIST, using the inputs from Figs. 2 and 3, are shown in Figure 4. Figure 4a shows the optical depth, ranging from less than 1 to up to 10 or 12 over the ocean. The cirrus cloud optical depths are very small, generally less than 1, but are difficult to visualize with this color scheme. the effective particle size is shown in Figure 4b. Most noticeable are the low r e of the stratus and the large r e of the cirrus. The water cloud re appear to be always less than 10 microns, hence more work will be done to investigate this feature. The ice clouds have re greater than 30 microns, therefore are not welldepicted in this figure. In Figures 4c and 4d, the accompanying SIST-derived liquid and ice water paths, respectively, are shown. In order to demonstrate the ability of this new application of SIST to larger areas in the GOES-R framework, a first full disk analysis was performed. In Figure 5 and Figure 6, a retrieval is shown for SEVIRI imagery from 1 August 2006 at 2300 UTC. The cloud optical depth and the the liqud water paths shown in the figures are promising when

a) c) b) d) Figure 4. Results from 2 August 2006 0200 UTC. SIST utilized the input from Figs. 3 and 4 to obtain a) optical depth, b) r e (microns), c) liquid water path (g/m 2 ), and d) ice water path (g/m 2 ). compared to similar retrievals of these quantities. 5. SUMMARY AND FUTURE WORK Results from this first application of a revised SIST are encouraging and show potential for application to full disk imagery from SEVIRI and ABI. While these results reveal some shortcomings, it is expected that these will be corrected as the algorithm matures. Potential enhancements to SIST, including the use of 8.7 and 13.3-µm data will be examined. Additionally, retrievals from day/night transition areas will be conducted to access the possibility of using SIST for twilight. Validation for both case studies and longer time scales remains a top priority, as well as comparisons to full-blown SIST algorithm retrievals from NASA Langley and from results provided by other GOES-R AWG team members. The fullblown SIST is used to process MODIS imagery data for the Clouds and the Earth Radiant Energy System (CERES) project, so MODIS-derived results that can also mimic ABI will be used. The integration of SIST into GEOCAT should facilitate

Fig. 5. SIST-derived cloud optical depth for 1 August 2006 2300 UTC. Fig. 6. SIST-derived cloud liquid water path (g/m 2 ) for 1 August 2006 2300 UTC. 6. REFERENCES Borbas, E., Moy, L., Seemann, S., Knuteson, R., Trigo, I., Antonelli, P., Li, J. and Huang, H. L., 2007: A global infrared land surface emissivity database and its validation. Proc. AMS 11 th Symposium on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface, 14-18. Hannon, S. E., L. L. Strow, and W. W. McMillan, 1996: Atmospheric infrared fast transmittance models: a comparison of two approaches, Proc. of SPIE 2830, 94 105. Kratz, D. P., 1995: The correlated k-distribution technique as applied to the AVHRR Channels, J. Quant. Spectroc. Radiat. Transfer, 53, 501-517. Minnis, P., D. P. Garber, D. F Young, R. F Arduini and Y. Takano, 1998: Parameterization of Reflectance and Effective Emittance for Satellite Remote Sensing of Cloud Properties. J. Atmos. Sci.,, 55, 3313-3339. Minnis, P., D. P. Kratz, J. A. Coakley, Jr., M. D. King, R. Arduini, D. P. Garber, P. W. Heck, S. Mayor, W. L. Smith, Jr., and D. F. Young, 1995: Cloud optical property retrieval (Subsystem 4.3). "Clouds and the Earth's Radiant Energy System (CERES) Algorithm Theoretical Basis Document, Volume III: Cloud Analyses and Radiance Inversions (Subsystem 4)", NASA RP 1376 Vol. 3, edited by CERES Science Team, December, pp. 135-176.