FIRMS users routinely leverage active fire detection data derived from MODIS and VIIRS. These sensors are onboard polar-orbiting satellites and provide a “snapshot” of fire activity at the time of satellite overpass for a given geographic area. Polar-orbiting observations by each individual MODIS and VIIRS sensor are conducted once to twice daily in the equatorial region of the globe and as many as eight times daily in the very high latitudes. What if you want multiple observations and updates on detected fire activity every hour? Geostationary satellites sensors can meet that need.

Unlike polar-orbiting satellite sensors, geostationary satellites sensors stay at a fixed point above the equator and follow the Earth’s daily rotation on its axis. Their orbits are 45-50 times higher than MODIS, VIIRS and Landsat sensors which operate in an orbit about 700 to 800 km above the Earth’s surface. These characteristics allow geostationary satellite sensors to move with the Earth’s rotation and persistently observe a very large portion of the Earth’s surface centered on their location.


GOES 16 (GOES East) coverage spans approximately 160o in both longitudinal and latitudinal directions.

GOES 16 (GOES East) coverage spans approximately 160o in both longitudinal and latitudinal directions.


Several geostationary platforms and sensors are designed to support meteorological observations and have spectral bands that span the visible, near infrared, and thermal infrared. These capabilities also enable them to support the detection and monitoring of fire activity. Additionally, geostationary sensors provide outstanding temporal resolution acquiring imagery for their entire field of view at 10-15 minute intervals or better. This enables geostationary sensors to potentially detect more fire events and capture their growth and change, particularly in between fire detection observations conducted by sensors on polar-orbiting platforms. However, geostationary sensors have a much coarser spatial resolution than polar-orbiting sensors and can be less sensitive to detecting relatively smaller fires.

Active fire detection data from five geostationary sensors are available in FIRMS and collectively provide global coverage. Two geostationary satellites, GOES-16 and GOES-18, are operated by NOAA and provide coverage for the western hemisphere. The other three satellites, Himawari-8, Meteosat 9 and Meteosat 11, are operated by JAXA and EUMETSAT and provide coverage for the eastern hemisphere and a substantial portion of the western hemisphere. Overlapping coverage is also provided by adjacent sensor fields of view. General information about the geostationary satellites used in FIRMS and their associated sensors, fire detection algorithms and derivative product information are summarized below.


Spatial coverage of geostationary satellites used in FIRMS (Ceamanos et al., 2021).

Spatial coverage of geostationary satellites used in FIRMS (Ceamanos et al., 2021).


Satellite

GOES-16, GOES-18

GOES-16, GOES-18

Meteosat 9 & 11

Himawari-8

Instrument/Algorithm

Advanced Baseline Imager (ABI) / Fire Detection and Characterization (FDC-HSC)

Advanced Baseline Imager (ABI) / Fire Radiative Power (FRP-PIXEL)

Spinning Enhanced Visible and Infra-Red Imager (SEVIRI) / Fire Radiative Power (FRP-PIXEL)

Advanced Himawari Imager (AHI) / Fire Radiative Power (FRP-PIXEL)

Satellite Source Agency

NOAA

NOAA

EUMETSAT

JAXA

Data Source

NOAA CLASS

Instituto Português do Mar e da Atmosfera (IPMA) under Copernicus Atmosphere Monitoring Service (CAMS)

EUMETSAT Land Surface Analysis Applications Facility (LSA SAF)

IPMA under Copernicus Atmosphere Monitoring Service (CAMS)

Coverage and Satellite Locations

Americas East and West

GOES-16: 0o, -75.2o

GOES-18: 0o, -137.2o

Americas East and West

GOES-16: 0o, -75.2o

GOES-18: 0o, -137.2o

Europe, Africa and Asia

Meteosat 9 (IODC): 0o, 45.5o, Meteosat 11: 0o, 0o

Asia and Australia

Himawari 8: 0o, 140o

More Information/Product User Manual (PUM)

NOAA GOES-R Series Advanced Baseline Imager (ABI) Level 2 Fire / Hot Spot Characterization (FDC) - Datasets

GOES FDC User Guide

FRP: GOES-R and Himawari Satellites PUM

See Xu et al. (2021)

Fire Radiative Power Pixel (FRPPIXEL, LSA-502)

See Wooster et al. (2015): See Roberts et al. (2015)

FRP: GOES-R and Himawari Satellites PUM

See Xu et al. (2021)

Algorithm

Filtered Fire Detection and Characterization (FDC-HSC) algorithm (provisional) (Schmidt et al., 2013). See: Why are the geostationary fire data filtered in FIRMS?

Geostationary Fire Thermal Anomaly (FTA) algorithm & FRP retrieval developed by King’s College, London. See Xu et al. (2021)

Geostationary Fire Thermal Anomaly (FTA) algorithm & FRP retrieval developed by King’s College, London. See Wooster et al. (2015)

Geostationary Fire Thermal Anomaly (FTA) algorithm & FRP retrieval developed by King’s College, London. See Xu et at. (2017)

Summary of geostationary satellites, sensors, algorithms and products used for active fire detection in FIRMS.



Here are two important considerations when using active fire detection data from geostationary satellites:  1) The coarse spatial resolution of geostationary satellites should be taken into consideration; the spatial resolution sub-nadir (i.e. directly below the satellite) is between 2-3Km for the different geostationary satellites (compared to 1Km, 375m, and 30m for MODIS, VIIRS, and Landsat respectively). An active fire could be located anywhere within that 2-3Km pixel. 2), Distortion in the pixel size of geostationary satellites increases towards the poles. This change in pixel size for the GOES-16 geostationary satellite is shown in the figure below. (To view the pixel sizes for the other geostationary satellites visit the FIRMS FAQ What is the spatial resolution of the geostationary satellite observations?)


Credit: Cullingworth and Mueller, 2021; https://doi.org/10.3390/rs13050878


Due to technical and environmental factors, active fire detection outputs from the current generation of geostationary algorithms can be prone to significant errors of commission and/or omission. Consequently, FIRMS considers all geostationary data as provisional or beta, and filters detections to display only those at the higher levels of detection confidence for each product. The filtered geostationary active fire detection data layers from each sensor/algorithm are available under the GEOSTATIONARY section of the Fire / Hotspots group in the FIRMS legend. They are accessible with the Advanced Mode selected. Each geostationary active fire detection layer is named by the source satellite and providing agency (see each layer’s information summary for more details). Additionally, for user convenience, the outputs for GOES FDC-HSC, Meteosat FRP-PIXEL and Himawari FRP-PIXEL algorithms are grouped into a single layer called Filtered Geostationary (provisional).

GOES ABI and Himawari AHI active fire detection data at 2km resolution are observed every 10 minutes throughout a 24-hour period while Meteosat SEVIRI active fire detection data at 3km resolution are observed every 15 minutes. This global harmonized, multi-sensor data stream is enabled by an automated framework developed by the University of Maryland under the auspices of the NASA Applied Sciences Program. All the sources of geostationary active fire detection data are available in FIRMS approximately 30 minutes or less post-observation. Users are advised to note that two active fire detection data products from GOES ABI imagery and derived from separate algorithms are provided in FIRMS.


Geostationary active fire detection layers in FIRMS


The current versions of geostationary active fire detection algorithms are undergoing additional development, refinement, and tuning. FIRMS will provide outputs based on new or updated algorithms when they are introduced into operational production by the source agencies. Stay tuned for additional FIRMS blog entries with more information on the algorithms, products and characteristics for geostationary active fire detection data and caveats users should consider when utilizing these data in FIRMS.

  • No labels