Elaine M. Prins NOAA/NESDIS/ORA/ARAD Advanced Satellite Products Team Madison, Wisconsin Joleen M. Feltz Chris C. Schmidt Cooperative Institute for Meteorological Satellite Studies Madison, Wisconsin Collection of Slides on Biomass Burning and Applications of Meteorological Satellite Fire Products

Applications of Meteorological Satellite Fire Products  Hazards Detection and Monitoring Each year millions of acres of forest and grassland are consumed by wildfire resulting in loss of life and property with significant economic costs and environmental implications. - Although the capabilities of current operational meteorological satellites are limited, they can provide valuable regional and global fire products in near real-time, and are critical for fire detection and monitoring in remote locations and developing countries.  Global Change Monitoring Biomass burning is a distinct biogeochemical process that plays an important role in terrestrial ecosystem processes and global climate change - Land use and land cover change monitoring: Fire is used in the process of deforestation and agricultural management. Approximately 85% of all fires occur in the equatorial and subtropical regions and are not adequately documented. - Estimates of atmospheric emissions: Biomass burning is a major source of trace gases and an abundant source of aerosols NO, CO 2 (40%), CO (32%), O 3 (38%), NO X, N 2 O, NH 3, SO X, CH 4 (10%), NMHC (>20%), POC (39%) - Within the Framework Convention on Climate Change (FCCC) countries will need to report on greenhouse gas emissions including those from biomass burning.

p 1-p Pixel How are Meteorological Satellites Used to Monitor Fires?

Examples of Automated Fire Detection Algorithms  Single channel thresholds e.g. AVHRR Instituto Nacional De Pesquisas Espaciais (INPE) fire product, European Space Agency ERS Along Track Scanning Radiometer (ATSR) fire product - Saturation in the 4 micron band - Elevated brightness temperature in the 4 micron band (I.e. > 315K)  Multi-channel thresholds e.g. Canada Centre for Remote Sensing (CCRS) Fire M3, CSU CIRA Fog/Reflectivity Product - 3 steps Use 4 micron band fixed thresholds to identify possible fires Use 11 micron band fixed thresholds to eliminate clouds Use 4 minus 11 micron band differences to distinguish fires from warm background  Contextual algorithms e.g. AVHRR Joint Research Centre of the European Commission (JRC) World Fire Web, Tropical Rainfall Mapping Mission (TRMM) Visible and Infrared Scanner (VIRS) GSFC fire product, AVHRR NOAA Fire Identification, Mapping and Monitoring Algorithm (FIMMA) fire product TERRA MODIS Fire Product - Implement multi-channel variable thresholds based on the heterogeneity of the background  Contextual identification and sub-pixel characterization e.g. UW-Madison GOES Automated Biomass Burning Algorithm(ABBA) - Implement contextual algorithms and determine estimates of sub-pixel fire size and temperature. Include offsets for emissivity and atmospheric attenuation.

Examples of Regional Fire Monitoring AVHRR Brazil INPE Fire Product June- November 2000 UW-Madison CIMSS GOES ABBA Fire Product June - October 1995 and 1997 NGDC DMSP Operational Linescan System (OLS) Fire Product, July - December 1997 Canada Centre for Remote Sensing (CCRS) Fire M3 Product 1999 fire season

GSFC TRMM Visible and Infrared Scanner (VIRS) Fire Product, July 2001 Examples of Global Fire Monitoring ESA ERS Along Track Scanning Radiometer (ATSR) Fire Product, August 1999 Joint Research Centre World Fire Web August 23 - September1, 2000

GSFC MODIS Fire Product November 1, 2000 Shenandoah National Park, VA Examples of Satellite Fire Monitoring in the United States NOAA AVHRR FIMMA Fire Product, August 23, 2001 UW-Madison CIMSS WFABBA July 11, 2001 Thirty Mile Fire, WA

Valley Complex, Bitteroot National Forest, MT July 31 – October 3, ,070 acres National Interagency Fire Center, Boise, Idaho Sula Complex, Sula, Montana August 6, 2000 John McColgan, BLM Alaska Fire Service NIFC MODIS, August 23, 2000, NASA GSFC GOES Composite for August 2000, UW-Madison/CIMSS 2000 Fire Season in the U.S. (NIFC) # of fires: 122, year average: 106,343 Acres burned: 8,422, year average: 3,786,411 Estimated Cost of Fire Suppression: $1.3 billion

Remote Sensing Fire Detection Validation Study for the 2000 Fire Season in Quebec When considering fires that burned more than 10 ha, the GOES and AVHRR were the first to detect many of the fires in the restricted protection zone of Quebec. Approximately 16 of the fires detected by the GOES were in remote locations and were not detected by the SOPFEU, Quebec’s forest fire detection and prevention agency.

Active Fire Detection in the Western Hemisphere Using the GOES Series

D C B A Arc of Deforestation Longitude °W Longitude °W Longitude °W Latitude °S Latitude °S Latitude °S Latitude °S Latitude °S Latitude °S Overview of Fires, Opaque Clouds, and Smoke/Aerosol Coverage in South America Derived from the GOES-8 ABBA and MACADA: FIRESSMOKE/AEROSOLOPAQUE CLOUDS

D C B A Longitude °W Longitude °W Latitude °S E J I H G F K L M N Latitude °S Latitude °S Fires Opaque CloudsSmoke/Aerosol Interannual Differences in Fires, Opaque Clouds, and Smoke/Aerosol: Each Fire Season (June - October) is Compared to the 1995 Benchmark Season

GOES-8 ABBA/MACADA South American Trend Analysis

Comparison of GOES-8 and AVHRR Fire Products for South America

Difference Plots of GOES-8 ABBA minus AVHRR INPE Fire Counts for South America Fire Seasons: 1995 through 2000

GOES South American ABBA Fire Products Used in Land Use/Land Cover Change and Fire Dynamics Research Universities, research institutes, and government planning agencies are using the GOES ABBA fire product as an indicator of land-use and land- cover change and carbon dynamics. GOES fire products also are being used to study the impact of road paving in South America on fire regime feedbacks and the future of the Amazon forests. Foster Brown, et al., 2001 Chico Mendes Extractive Reserve

MOPITT CO composite is courtesy of the MOPITT team: John Gille (NCAR), James Drummond (University of Toronto), David Edwards (NCAR) EOS MOPITT identifies elevated carbon monoxide associated with biomass burning detected with the GOES ABBA GOES-8 South American ABBA Composite Fire Product September 7, 2000 Comparison of GOES ABBA Fire Observations and the EOS MOPITT CO Product

GOES-8 Experimental AOT Analysis: 26-August-1995 at 1445 UTC Panel a shows smoke as observed in GOES-8 visible imagery at 1445 UTC on 26 August The GOES-8 MACADA is applied to multi- spectral GOES imagery to automatically distinguish between clear-sky, cloud type, and smoke using both textural and spectral techniques (panels b and c). The MACADA textural/spectral analysis is then used to create a 14-day clear-sky background reflectance map shown as solar zenith weighted albedo (panel d). The AOT product (panel f) is determined by calculating the difference between the observed and clear-sky reflectance and relating this to AOT utilizing a look-up table. If no clear-sky reflectance is available for a given location, an ecosystem based reflectance map (panel e) is used. a. b. c. d. e. f.

GOES-8 Derived AOT During Study Period: 8/15/95 - 9/7/95 at 1445 UTC Using Single Scattering Albedos.83,.90, and.95 Single Scattering Albedo: 0.83Single Scattering Albedo: 0.90Single Scattering Albedo: 0.95 r = 0.74 r = 0.93 r = 0.89

GOES-8 Derived AOT for North Atlantic Basin Study Period: July 11-31, 1996 at 1345, 1445, 1545, 1645, 1745, 1845 UTC

Animations of Wildfire ABBA composite image products are being provided via anonymous ftp and the web every half-hour at: Displays include three overviews and 35 regional views providing coverage of the entire Western Hemisphere. Examples of Regional View Sectors GOES Wildfire ABBA Fire Composite Web Distribution

Examples of the GOES Wildfire ABBA Monitoring System in the Western Hemisphere

GOES-8 Wildfire ABBA Summary Composite of Half-Hourly Temporally Filtered Fire Observations for the Western Hemisphere Time Period: September 1, 2000 to August 31, 2001 The composite shows the much higher incidence of burning in Central and South America, primarily associated with deforestation and agricultural management. Fire Pixel Distribution 30-70°N: 6% 10-30°N: 10% 70°S-10°N: 84% Wildfires Agricultural Burning Crops and Pasture Recent Road Construction Desert/Grassland Border Deforestation/maintenance

Model Data Assimilation Activities - At the Naval Research Laboratory (NRL-Monterey) GOES ABBA fire product information is being assimilated into the Navy Aerosol Analysis and Prediction System (NAAPS) to analyze and predict aerosol loading and transport as part of the NASA-ESE Fire Locating And Mapping of Burning Emissions (FLAMBE) project. - Model output is being compared to GOES satellite derived aerosol products and TOMS products. Initial studies show the model output and aerosol products are in close agreement.

GOES-8 Wildfire ABBA fire product for the Pacific Northwest Date: August 17, 2001 Time: 2200 UTC NAAPS Model Aerosol Analysis for the continental U.S. Date: August 18, 2001 Time: 1200 UTC Real-Time Aerosol Transport Model Assimilation of the Wildfire ABBA Fire Product FIRES

Future Environmental Satellite Fire Monitoring Capabilities  Global Geostationary Fire Monitoring System - GOES-E/W Imager - METEOSAT Second Generation (MSG) (2002) Spinning Enhanced Visible and InfraRed Imager (SEVIRI) - Multi-functional Transport Satellite (MTSAT-1R) (2003) Japanese Advanced Meteorological Imager (JAMI)  NOAA Operational Systems - NPOESS Preparatory Project Visible/Infrared Imager Radiometer Suite (VIIRS) (2006) - Advanced Baseline Imager (ABI) (2010)  International Platforms Designed for Fire Detection - German Aerospace Center (DLR) Bi-spectral Infrared Detection (BIRD) - German Aerospace Center (DLR) Intelligent Infrared Sensor System (FOCUS) (ISS) - Consortium of DLR and European space industries are designing the Forest Fire Earth Watch (FFEW-FUEGO) satellite mission

GOES-EGOES-WMSGMTSAT Satellite View Angle 80° 65° International Global Geostationary Active Fire Monitoring: Geographical Coverage 322 ha

Minimum Detectable Fire Size Estimates for GOES, MSG, and MTSAT

GOES-R and GOES-I/M Simulations of Viejas Fire Using MODIS Data: January 3, 2001 at 1900 UTC Simulated GOES-R: 3.9 micronSimulated GOES-I/M: 3.9 micron

GOES-R and GOES-I/M Simulations of Viejas Fire Smoke Plume Using MODIS Data Simulated GOES-R: visible:.64 micron January 3, UTC Simulated GOES-I/M: visible:.64 micron January 3, UTC