Active Fire Monitoring: Product Guide



Similar documents
Volcanic Ash Monitoring: Product Guide

EUMETSAT Satellite Programmes

The impact of window size on AMV

MSG MPEF Products focus on GII Simon Elliott Meteorological Operations Division

The APOLLO cloud product statistics Web service

Synoptic assessment of AMV errors

Evaluation of Wildfire Duration Time Over Asia using MTSAT and MODIS

Best practices for RGB compositing of multi-spectral imagery

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

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

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

Overview of the IR channels and their applications

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

Cloud Masking and Cloud Products

Sentinel-3 Marine Test Data Set processed at EUMETSAT

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

A comparison of NOAA/AVHRR derived cloud amount with MODIS and surface observation

THE USE OF THE HIGH RESOLUTION VISIBLE IN SAFNWC/MSG CLOUD MASK

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

Options for filling the LEO-GEO AMV Coverage Gap Francis Warrick Met Office, UK

Monitoring of Arctic Conditions from a Virtual Constellation of Synthetic Aperture Radar Satellites

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

Emergency Management Service. early warning FLOOD AND FIRE ALERTS. Space

Outline of RGB Composite Imagery

Estimating Firn Emissivity, from 1994 to1998, at the Ski Hi Automatic Weather Station on the West Antarctic Ice Sheet Using Passive Microwave Data

GLOBAL FORUM London, October 24 & 25, 2012

Review for Introduction to Remote Sensing: Science Concepts and Technology

Satellite Remote Sensing of Volcanic Ash

Daily High-resolution Blended Analyses for Sea Surface Temperature

MSG-SEVIRI cloud physical properties for model evaluations

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

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

INVESTIGA I+D+i 2013/2014

A remote sensing instrument collects information about an object or phenomenon within the

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

STAR Algorithm and Data Products (ADP) Beta Review. Suomi NPP Surface Reflectance IP ARP Product

Analysis of Climatic and Environmental Changes Using CLEARS Web-GIS Information-Computational System: Siberia Case Study

MODIS IMAGES RESTORATION FOR VNIR BANDS ON FIRE SMOKE AFFECTED AREA

McIDAS-V Tutorial Displaying Polar Satellite Imagery updated September 2015 (software version 1.5)

Satellite Snow Monitoring Activities Project CRYOLAND

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

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

Product Navigator User Guide

Climatology and Monitoring of Dust and Sand Storms in the Arabian Peninsula

How To Understand Cloud Properties From Satellite Imagery

Present Status of Coastal Environmental Monitoring in Korean Waters. Using Remote Sensing Data

Information Contents of High Resolution Satellite Images

COMPUTING CLOUD MOTION USING A CORRELATION RELAXATION ALGORITHM Improving Estimation by Exploiting Problem Knowledge Q. X. WU

Land Use/Land Cover Map of the Central Facility of ARM in the Southern Great Plains Site Using DOE s Multi-Spectral Thermal Imager Satellite Images

CLOUD FREE MOSAIC IMAGES

VIIRS-CrIS mapping. NWP SAF AAPP VIIRS-CrIS Mapping

TerraAmazon - The Amazon Deforestation Monitoring System - Karine Reis Ferreira

Data Processing Flow Chart

High Resolution Information from Seven Years of ASTER Data

An Introduction to the MTG-IRS Mission

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

Vulnerability assessment of ecosystem services for climate change impacts and adaptation (VACCIA)

Precipitation Remote Sensing

Nowcasting of significant convection by application of cloud tracking algorithm to satellite and radar images

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

Integrated Global Carbon Observations. Beverly Law Prof. Global Change Forest Science Science Chair, AmeriFlux Network Oregon State University

Forest fire detection for near real-time monitoring using geostationary satellites

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

GOES-R AWG Cloud Team: ABI Cloud Height

Land-surface emissivity maps based on MSG/SEVIRI information

Supervised Classification workflow in ENVI 4.8 using WorldView-2 imagery

CGMS-36, NOAA-WP-14 Prepared by C. Cao Agenda Item: II/2 Discussed in WG II

SAMPLE MIDTERM QUESTIONS

METEOSAT. Europe s geostationary

Forest Fire Information System (EFFIS): Rapid Damage Assessment

NCDC s SATELLITE DATA, PRODUCTS, and SERVICES

How To Find Natural Oil Seepage In The Dreki Area Using An Envisat Image

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

GCOS science conference, 2 Mar. 2016, Amsterdam. Japan Meteorological Agency (JMA)

Generation of Cloud-free Imagery Using Landsat-8

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

McIDAS-X Software Development and Demonstration

Eastern Caribbean Open-Source Geospatial Data Sharing and Management Workshop

Night Microphysics RGB Nephanalysis in night time

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

A KNOWLEDGE-BASED APPROACH FOR REDUCING CLOUD AND SHADOW ABSTRACT

Transcription:

Active Fire Monitoring: Product Guide Doc.No. Issue : : EUM/TSS/MAN/15/801989 v1c EUMETSAT Eumetsat-Allee 1, D-64295 Darmstadt, Germany Tel: +49 6151 807-7 Fax: +49 6151 807 555 Date : 14 April 2015 http://www.eumetsat.int WBS/DBS : EUMETSAT The copyright of this document is the property of EUMETSAT.

Document Change Record Issue / Revision Date DCN. No Summary of Changes 1 06/10/2010 Initial creation of document in DM tool. 1A 08/11/2010 Initial content and graphics added. 1B 18 Nov 2014 Added content to specify product output and specifications to fit Product Guide template. 1C 14 April 2015 Subject matter expert review. Section 4 updated with algorithm specifications for numerical indicators and updated scene quality information. Table of Contents 1 PRODUCT DESCRIPTION... 3 2 PRODUCT SPECIFICATIONS... 3 2.1 Product history:... 4 3 PRODUCT ILLUSTRATION... 5 3.1 Analysis of results after improvements to the algorithm in 2008... 7 3.2 Major product improvement made after product validation... 7 3.3 Numerical Weather Prediction (NWP) number assignments... 7 4 BASIC STRUCTURE OF THE ALGORITHM... 10 4.1 Input Data Used... 10 4.2 Algorithm Description... 10 5 REFERENCE AND LINKS... 11 5.1 Reference Documents... 11 5.2 Online Resources and Assistance... 11 Page 2 of 11

1 PRODUCT DESCRIPTION The Active Fire Monitoring (FIR) product is an image-based meteorological product in full pixel resolution that displays information on the presence of fire within a pixel. It was developed primarily for detecting and quantifying biomass burning in Africa, but is generated for the complete MSG footprint. It is generated for every image and compares the brightness temperatures of the IR3.9, 8.7 and 10.8 channels over cloud-free land areas, then assigns one of three fire statuses to each pixel: fire-free, potential, or probable. The algorithm is based on threshold values but there are plans for further development. The underlying concept of the algorithm takes advantage of the fact that SEVIRI channel IR3.9 is very sensitive to hot spots which are caused by fires. The algorithm distinguishes between potential fire and active fire. To classify a pixel as active fire, the algorithm uses a stricter set of classification criteria than for potential fires. Classifications of potential fires are pixels with small fires, or pixels with burned-out fires but still-hot surfaces This category can even include some false classifications possibly caused by other hot surfaces, regions with inhomogeneous surfaces, or with undetected small clouds. In Europe, these fires occur mainly during summer and early autumn. In sub-saharan Africa, however, fires of this magnitude occur throughout the year. 2 PRODUCT SPECIFICATIONS Category Type Product abbreviation Applications and users Input satellite data Product Distribution Product Area Product Resolution Meteorological Product FIR Specification fire detection, early warning for fire potential MSG images from the IR10.8 channel and passive microwave data from the SSM/I instrument on the US Defense Meteorological Satellite Program spacecraft. EUMETCast EUMETSAT Data Centre Full earth scanning (FES) area RSS Area Pixel Page 3 of 11

Category Product Distribution Frequency Full Earth Scanning Area Specification EUMETCast: every 15 minutes for the 00:00, 00:15, 00:30, 23:45 UTC products EUMETSAT Data Centre: every 15 minutes for the 00:00, 00:15, 00:30, 23:45 UTC products Product Format Product Size Resolution Rapid Scanning Service Area EUMETCast: every 5 minutes for the 00:00, 00:05, 00:10, 23:55 UTC products EUMETSAT Data Centre: every 5 minutes for the 00:00, 00:05, 00:10, 23:55 UTC products GRIB2 format (FIRG product) CAP format (Common Alert Protocol) Full Earth Scanning Area: GRIB product 20 kb (variable) CAP product 5 kb (variable) Rapid Scanning Service Area: GRIB product 4 kb (variable) CAP product 2 kb (variable) 3 km 3 km pixel resolution 2.1 Product history: Initial operational dissemination: Improvements to product after validation testing Changes to product inputs: Replacement of ASCII text file with CAP product. Corrective maintenance Substantial gaps in coverage 2007 Dissemination via EUMETCast in ASCII format. 2008 Significantly improved hot-spot detection 28 April 2009 New surface emissivity maps based on MODIS observation added to product processing. Dynamic thresholds from ECMWF used to replace static thresholds. See Section Error! Reference source not found. for a detailed description. 24 Mar 2011 The new CAP-formatted product was disseminated on the alert channel on EUMETCast. 8 Feb 2012 Corrective maintenance on FIR-DU and pixel radius in the CAP-formatted product. None Page 4 of 11

3 PRODUCT ILLUSTRATION Product results are shown in comparison images in Figure 1 and Figure 2 that follow. Figure 1 shows an IR3.9 image for 21 August 2005 at 02:00 UTC over Portugal and western Spain both before and after application of the product. The location of the hot spots either active fires or potential fires is clearly marked in the lower image when the Active Fire Monitoring product is applied. Red pixels are active fires, yellow pixels are potential fires. Figure 2 also shows the IR3.9 image before and after the Active Fire Monitoring product has been applied. The image is from south central Africa near Lake Tanganyika and Lake Malawi. Again, red pixels are active fires, yellow pixels are potential fires. Figure 1: Top panel: Initial IR3.9 image for 21 August 2005 at 02:00 UTC. Bottom panel: Active Fire Monitoring product applied to IR3.9 image. Page 5 of 11

Figure 2: Initial IR3.9 image for 21 August 2005 at 12:00 UTC over central Africa is at left. Active Fire Monitoring product is applied to the image at right. Page 6 of 11

3.1 Analysis of results after algorithm improvements in 2008 The first analysis was the inspection of individual products to demonstrate the expected improvement in the hot spot detection. The example presented here is from 18 August 2008 11:30 UTC using Meteosat-9 full Earth scan (FES) products. Figure 3 shows the underlying IR3.9 μm channel image over the west coast of Africa. Figure 4 shows the image after application of the FIR product. On the left is the operational algorithm. The new FIR algorithm is on the right. Red pixels indicate probable fires, yellow indicates possible fires. A first look at the images in Figure 4 reveals that the operational algorithm misses many of the fires that exist, whereas the new algorithm (at right) manages to capture the hot spots in the IR3.9 μm imagery very well. In the non-cloudy area in the eastern part of the image, the operational FIR product misses a large number of hot spots. Even where the signal is weaker (as in the southern part of the area) the detection rate has improved. However, it should be noted that, occasionally, the improved algorithm does not detect a fire that has been detected by the old operational algorithm, as shown in the south-western corner of the plot. Nevertheless, the detection rate has clearly improved. The detection of fires in the cloudy area has also improved, but only slightly. The latter analysis was supported by several case studies performed on some of the undetected hot spots in the previous operational product. These studies showed that often these fires were not detected because the IR10.8 μm standard deviation test has a strict threshold of less than 1 K. This means that the standard deviation of IR10.8 μm brightness temperature needs to be lower than 1 K within a 3 x 3 pixel image segment in order to allow fire detection. The purpose of these studies was to remove false alarms for fires that were sometimes detected along coastlines, rivers and around cloud edges due to these parameters. 3.2 Major product improvement made after product validation During operations product validation, two major changes were made to the Active Fire Monitoring (FIR) product. The first change concerns the use of new surface emissivity maps. In the previous product release, the emissivity was set to 1.0 over land. The algorithm was changed to use monthly averages of surface emissivity data derived from MODIS observations instead. The second change concerns the thresholding technique used. In the previous version, static thresholds were used, whereas in the improved version dynamic thresholds derived from ECMWF forecast data are used. With this improved functionality, the product can better cope with area differences in both time and space. 3.3 GRIB-2 Encoded profuct number assignments The GRIB-2 encoded product produces these values for each pixel. Values are in accordance with WMO FM 92-XIII GRIB Code Table 4.223 Fire detection indicator. Number...means 0 No fire detected. 1 Possible fire detected 2 Probable fire detected 3 Missing Page 7 of 11

Figure 3: Basic IR3.9 channel image corresponding to FIR images in Figure 4. Page 8 of 11

Figure 4: (Left) FIR product over western Africa for 18 August 2008 11:30 UTC, using the operational algorithm. Red indicates a probable fire, while yellow indicates a possible fire. (Right ) The FIR product over western Africa for 18 August 2008 11:30 UTC, using the improved algorithm. Red is a probable fire, while yellow is a possible fire. Page 9 of 11

4 BASIC STRUCTURE OF THE ALGORITHM 4.1 Input Data Used MSG SEVIRI image data in the following channels: VIS0.6 µm channel IR3.9 µm channel IR10.8 µm channel Forecast data from the European Centre for Medium Range Weather Forecasts (ECMWF) which are used for estimating the Equivalent Blackbody Brightness Temperature (EBBT) in the IR channels. Results of the pixel-based Scenes Analysis (an internal product), indicating for every pixel whether cloud has been detected Surface emissivity, coastline and surface type maps 4.2 Algorithm Description The basic principles of the FIR algorithm are similar to those already in use for other instruments like GOES, AVHRR and MODIS. Scene quality is taken from the pixel-based Scenes Analysis product. The scale in this product is as follows: 100 means that the pixel is cloudy, while 0 means the pixel is clear. Given this criteria, off-shore oil fires, fires from active volcanoes on small islands, or fires hidden by thick clouds are not monitored by the algorithm. Bare-soil land surfaces (deserts) are also excluded from the processing. During daylight hours, channel VIS0.6 is used to exclude bare soil surfaces, small clouds, and sun glint areas. For the remaining valid pixels, the FIR algorithm uses the following criteria to check for potential fire and fire pixels: Brightness temperature of channel IR3.9 Standard deviation of channel IR3.9 Brightness temperature difference of channel IR3.9 and IR10.8 Standard deviation of channel IR10.8 4.3 Known Operational Limitations The status of the product is demonstration. Page 10 of 11

5 REFERENCES AND LINKS 5.1 Reference Documents Validation Type Document Name Reference Operations Product Validation Report: FIR Product EUM/OPS/REP/08/3689 Detailed Algorithm Science/Research Science/Research MSG Meteorological Products Extraction Facility Algorithm Specification Document Amraoui, M., C.C. DaCamara, J.M.C. Pereira, 2010: Detection and monitoring of African vegetation fires using MSG-SEVIRI imagery Giglio L., J. Descloitres, C.O. Justice, Y.J. Kaufman 2003: An Enhanced Contextual Fire Detection Algorithm for Modis. EUM/MSG/SPE/022 doi:10.1016/j.rse.2009.12.019 Remote Sensing of Environment, Vol. 87, pp. 273-282 5.2 Online Resources and Assistance All of the reference documents listed above are in the EUMETSAT Technical Documents page. www.eumetsat.int > Satellites > Technical Documents > Meteosat Services > 0 Meteosat Meteorological Products A training presentation for the FIR Product is here: http://www.eumetsat.int > home > Data > Training > TrainingLibrary > Index > Use of SEVIRI and AVHRR Channels for Remote Fire/Smoke Detection To register for data delivery from this product, go to the Data Registration page on the EUMETSAT web page: www.eumetsat.int > Data > Data Delivery > Data Registration Information about the service status of EUMETSAT satellites and the data they deliver is this EUMETSAT web page: www.eumetsat.int > Data > Service Status To get answers to any questions about data delivery, registration or documentation, contact the EUMETSAT User Service Help Desk: Telephone: +49 6151 807 3660/3770 e-mail: ops@eumetsat.int Page 11 of 11