Monitoring Fires from Space and Getting Data in to the hands of Users
Presented by Diane Davies1,2,
Operations Manager for NASA LANCE (Land Atmosphere Near real-time Capability for EOS)
1 Trigg-Davies Consulting Ltd, 2 Sigma Space Corporation
ZSL Symposium on Remote Sensing for Conservations: Uses, Prospects and Challenges Fires in Irkutsk region, Russia
MODIS Image from Aqua satellite acquired May 18th, 2014
Earth Observatory Image of the Day http://1.usa.gov/TpaD0J
Fires in Southern California and Mexico
MODIS Image from Aqua satellite acquired May 14th, 2014
Source: NASA Earth Observatory Image of the Day http://1.usa.gov/1k1BEkL
Fires in Irkutsk region, Russia
MODIS on Terra satellite captured May 18th, 2014
Earth Observatory Image of the Day http://1.usa.gov/TpaD0 J
MODIS image acquired from the Aqua satellite on May 8, 2014 Source: NASA Earth Observatory Image of the Day http://1.usa.gov/1oablsK The MODIS sensor
MODIS (Moderate Resolution Imaging Spectroradiometer) launched on NASA’s Terra Earth Observing System (EOS) satellite in December 1999, providing morning and night time high quality observations since 2000. MODIS launched again on NASA’s Aqua EOS satellite in 2002.
MODIS global products were developed by science teams to serve the needs of the broader user community.
2012 MODIS Active Fire Detections
Prior to MODIS
NOAA AVHRR First launched in 1978 to provide daily global coverage on clouds, snow and ice and surface temperature After launch RS community realized it could be used to detect fires No one algorithm performed optimally over all biomes Efforts to get data to users included direct broadcast stations Direct broadcast of NOAA AVHRR
Etosha National Park, Namibia
System provided by LARST - Local applications of Remote Sensing Technology with funds from the UK ODA (now Department for International Development (DfID). MODIS Global Browse landweb.nascom.nasa.gov
Daily Daytime Active Fire Detection over Daily Land Surface Reflectance (MYD14 over MYD09)
May 5, 2014 Deer and wildfire, Montana Taken 6 August 2000 by John McColgan, USFS, via Wikimedia Commons. MODIS Rapid Response
Source: Sohlberg, R., J. Descloitres, et al. (2001). "MODIS Land Rapid Response: operational use of Terra data for USFS wildfire management." The Earth Observer 13(5): 8-10,14. Interface for the Namibia Web Fire Mapper Service showing fires outlined in red and locally supplied GIS layers superimposed on the most recently acquired MODIS image. Fire Information for Resource Management System
2006 FIRMS was funded under NASA Applied Sciences Program
- UMD in partnership with UN FAO and Conservation International - Further develop prototype - Assist Conservation International further develop their email alert service -- Establish operational system at FAO
FIRMS Fire Email Alert 13
Fire Email Alerts
“When FIRMS gives us evidence of a fire and its exact coordinates, that is very valuable information.”
Rafael Manzanero Friends for Conservation and Development
Photo courtesy of Friends for Resource manager Rafael Manzanero in the Chiquibul National Park, Belize. Conservation and Development, Belize) (Photograph by J. Houston courtesy Rare)
Article “Orbiting watchtowers” by Natasha Vicarra, Sensing our Planet: NASA Earth Science Research Features 2012 http://earthdata.nasa.gov/sensing -our-planet Global Fire Information Management System http://firecast.conservation.org
Advanced Fire Information System (AFIS) for South Africa Useful but…
• Not all fires are detected • Each hotspot represents the center of a 1km (approx.) pixel
Collection 6 reprocessing is underway • Processing extended to oceans and other large water bodies - Detect off-shore gas flaring • Reduce false alarms in Amazon caused by small forest clearings • Dynamically adjust potential fire thresholds • Detect smaller fires • Improved cloud mask • Improved detection confidence estimate
FAQs https://earthdata.nasa.gov/data/near-real-time-data/faq/firms Australia 1km active fires 1 month 2002
David Roy, SDSU Australia 500m burned areas 1 month 2002
Color Code Gives the Timing of the Burn within the Month David Roy, SDSU MODIS Burned Area MCD45A1
For more information: http://modis-fire.umd.edu/Burned_Area_Products.html Land Atmosphere Near real-time Capability for EOS (LANCE) https://earthdata.nasa.gov/lance LANCE makes EOS data from MODIS, AIRS, MLS and OMI available within three hours of satellite overpass to meet the timely needs of applications such as numerical weather and climate prediction; forecasting and monitoring natural hazards, ecological/invasive species, agriculture, and air quality; providing help with disaster relief; and homeland security.
MODIS Rapid Response Subsets
MODIS / Terra Corrected Reflectance bands 7,2,1, with active fires overlaid in red 9th May 2014, Northern Australia Example from Kruger National Park: using MODIS subset images to improve ranger estimates of area of burn
Source: Navashni Govender, KNP https://earthdata.nasa.gov/worldview
EOSDIS Global Imagery Browse Services (GIBS) and Worldview
Driving Goal: To transform how end users interact with and discover EOSDIS data; make it visual
Approach: Open
- – The Global Imagery Browse Services (GIBS) ServersAccess provide access to full resolution imagery derived from NASA products in a standardized
manner to any web-connected client
– Worldview provides a highly responsive interface to explore GIBS imagery in a Google Client
Maps-like manner and download the underlying data
Global Imagery Browse Services / Worldview https://earthdata.nasa.gov/gibs https://earthdata.nasa.gov/worldview
• Global Imagery Browse Services (GIBS) • Serves global, daily, full resolution imagery for 100+ MODIS, AIRS, OMI, MLS products • Most imagery is updated within three hours of acquisition • Free and open access via standardized protocols • Highly responsive for interactive client access (e.g., OpenLayers, Google Earth, Leaflet) and script-level access via GDAL • Geographic, Arctic, and Antarctic map projections available; Web Mercator coming soon • Daily imagery is available from May 2012 to present; backfill of MODIS data from 2000- present currently underway
• Worldview • Web browser-based client to interactively explore full-resolution imagery served by GIBS • Ability to download imagery and underlying data • Compatible with desktop and mobile browsers • Results are shareable via URL-based “permalinks”
International Global Geostationary Active Fire Monitoring: Geographical Coverage
80 -160 -120 -80 -40 0 40 80 120 160 GOES-W GOES-E MSG MTSAT 60
40 Satellite View Angle 20 80° 65° 0 -20
-40
-60
Wildfire Automated Biomass Burning Algorithm (WFABBA) uses geostationary data to detect biomass burning (UW-Madison) http://wfabba.ssec.wisc.edu MODIS Global Browse landweb.nascom.nasa.gov
Daily Daytime Active Fire Detection over Daily Land Surface Reflectance (MYD14 over MYD09)
May 5, 2014 Suomi NPP VIIRS Global Browse landweb.nascom.nasa.gov
Daily Daytime Active Fire Detection over Daily Land Surface Reflectance (MYD14 over MYD09)
September 21, 2012 Slide courtesy of Chris Justice, UMD NASA LandPEATE MODIS and VIIRS fire detections at nadir: modeling VIIRS spatial resolution is higher that of MODIS; in general, VIIRS is expected to detect smaller fires at nadir (based on modeling using ASTER fire masks)
MODIS VIIRS (aggregated)
7 Aug 2004 1405 UTC ~11.7o S 56.6o W (Brazil) Slide courtesy of Chris Justice, UMD VIIRS Active Fire Product Website
viirsfire.geog.umd.edu Acknowledgements
• MODIS Active Fire: Chris Justice1, Louis Giglio1 • MODIS Burned Area: David Roy2, Luigi Boschetti3 • VIIRS: Ivan Csiszar4, Wildfrid Schroeder1, 4, Louis Giglio1, Evan Ellicot1, Chris Justice1 and Krishna Vedrevu1 LANCE: Kevin Murphy5, Karen Michael5, Dawn Lowe5 Worldview: Ryan Boller5, Jeff Schmaltz6, Mike McGann7, Taylor Gunnoe7, Shriram Ilavajhala6, GIBS: Matt Cechini7, Mike McGann7, Jeff Schmaltz6, Shriram Ilavajhala6 GOFC-FIRE: Krishna Vedrevu1, Chris Justice1 NASA Earth Observatory: Adam Voiland6, Jesse Allan6 FIRMS: Diane Davies6, Minnie Wong7, Shriram Ilavajhala6 GFIMS: John Latham8, Antonio Martucci8 Conservation International Firecast: Karyn Tabor9 AFIS: Philip Frost10
1Univ. of Maryland, 2Univ. of South Dakota, 3Univ. of Idaho, 4NOAA, 5NASA GSFC, 6Sigma Space Corp., 7Columbus Technologies & Services, 8UN FAO, 9Conservation International, 10CSIR Meraka
GOFC/GOLD - Fire
GOFC-Fire Implementation team of GOFC /GOLD - Global Observation of Forest and Land Cover Dynamics. A panel of the Global Terrestrial Observing System (GTOS)
• The main goal of GOFC-GOLD is to provide a forum for international information exchange, observation and data coordination and a framework for establishing the necessary long- term monitoring systems.
• Increase user awareness by improved understanding of the utility of satellite fire products for resource management
• Advocating for Data Continuity and new and better Missions
• Standards and Protocols for Fire Data Products
Contact: Krishna Vadrevu, University of Maryland
What fire information is needed for conservation?
Active Fire detection* Burned area mapping* Fire danger rating Fire intensity Estimate of biomass burned Emissions for REDD+
*global products Key factors in the uptake of MODIS active fire data
User confidence in a product is key to uptake (refinement of algorithm and validation of products using ASTER) End user feedback Information not data Small, easy to use formats User friendly service Educating users about the pitfalls of using these data is key Importance of context and imagery Aqua MODIS vs. Suomi NPP VIIRS • Aqua and NPP have similar overpass times (1:30pm) – sampling of the diurnal fire cycle is similar • Saturation levels of the primary bands allow unsaturated radiance measurements for most fires – Bands 21/22 for MODIS and M13 for VIIRS • Some differences in spectral placement • Processing algorithms are compatible – Current VIIRS algorithm is based on MODIS, albeit an earlier version – Differences can be resolved and the impact can be minimized • Primary driver of differences is spatial sampling – Pixel size – Variations along scanline (aggregation schemes) – Variations within pixels (line-spread function, aggregation) – Differences in swath width (VIIRS has no gaps at low latitudes) LANCE vs. Standard Product Latency – MODIS Example Standard Processing LANCE Processing (typical) Product Category Terra(hrs) Aqua(hrs) Terra/Aqua (hrs)
L1/Cloud Mask 8 25 1.7
L2 Snow 8 25 1.8
L2 Sea Ice 8 25 2.0
L2 Fire 8 25 1.9
L2 Clouds 32 32 2.2
L2 Aerosol 32 32 2.2
L2 LSR 40 41 2.1
42 Near Real-Time vs. Science Quality Products – MODIS
Science ProductExample Near Real -Time Product
Land Surface Reflectance Reflectance Land Surface
Cloud Top Temperature Cloud Top
43 MODIS Fire User Guide (v. 2.5): select FAQs
How are the fires and other thermal anomalies identified in the MODIS fire products detected? Fire detection is performed using a contextual algorithm (Giglio et al., 2003) that exploits the strong emission of mid-infrared radiation from fires (Dozier, 1981; Matson and Dozier, 1981). The algorithm examines each pixel of the MODIS swath, and ultimately assigns to each one of the following classes: missing data, cloud, water, non-fire, fire, or unknown. Pixels lacking valid data are immediately classified as missing data and excluded from further consideration. Cloud and water pixels are identified using cloud and water masks, and are assigned the classes cloud and water, respectively. Processing continues on the remaining clear land pixels. A preliminary classification is used to eliminate obvious non-fire pixels. For those potential fire pixels that remain, an attempt is made to use the neighboring pixels to estimate the radiometric signal of the potential fire pixel in the absence of fire. Valid neighboring pixels in a window centered on the potential fire pixel are identified and are used to estimate a background value. If the background characterization was successful, a series of contextual threshold tests are used to perform a relative fire detection. These look for the characteristic signature of an active fire in which both 4 μm brightness temperature and the 4 and 11 μm brightness temperature difference depart substantially from that of the non-fire background. Relative thresholds are adjusted based on the natural variability of the background. Additional specialized tests are used to eliminate false detections caused by sun glint, desert boundaries, and errors in the water mask. Candidate fire pixels that are not rejected in the course of applying these tests are assigned a class of fire. Pixels for which the background characterization could not be performed, i.e. those having an insufficient number of valid pixels, are assigned a class of unknown.
What is the smallest fire size that can be detected with MODIS? What about the largest? MODIS can routinely detect both flaming and smoldering fires ∼1000 m2 in size. Under very good observing conditions (e.g. near nadir, little or no smoke, relatively homogeneous land surface, etc.) flaming fires one tenth this size can be detected. Under pristine (and extremely rare) observing conditions even smaller flaming fires ∼50 m2 can be detected. Unlike most contextual fire detection algorithms designed for satellite sensors that were never intended for fire monitoring (e.g. AVHRR, VIRS, ATSR), there is no upper limit to the largest and/or hottest fire that can be detected with MODIS.