Active Fires - Total Operational Weather Readiness - Satellites (TOWR-S)
VIIRS Active Fires
Frequently Asked Questions about the VIIRS Active Fires Product
1) What is this product?
This product displays points where its algorithm believes a fire may be occuring. It also has confidence percentages for each point, which represent how confident the algorithm is that the fire is indeed an actual fire. Finally, the product produces a radiative power in MW for each fire pixel. This product is useful when trying to identify the exact location of an ongoing fire, or the location of a new fire that has not yet been observed from the ground (such as a fire sparked by lightning during a thunderstorm). The confidence intervals can assist a forecaster in assessing the likelihood that the algorithm is indeed picking up on a fire, or if it is a false detection. The radiative power of the fire can be used to assess the general strength/temperature of the fire.
All of this information can be used to relay information to emergency managers about new fires, or be used by Incident Meteorologists (IMETs) already deployed to a fire to relay the location of the fire's leading edge, in addition to assist with forecasting of where the fire is likely to move next. Because of the 375m resolution of this product, it is the highest-resolution fire product available to the NWS forecaster. Note that pixel saturation may be found over fires typically spreading over multiple contiguous pixels, an indication of potentially large and intense biomass burning. 1
2) How often do I receive this data?
The S-NPP satellite is part of the Afternoon Train (A-Train) of satellites. It crosses your area at ~1:30am and ~1:30pm local time every day.
3) How do I display this product in AWIPS-II?
4) How do I interpret the color maps associated with this product?
5) What should I use in conjunction with this product to produce a better forecast?
Geostationary SWIR imagery can sometimes view fires as well, and can act as a secondary source of information to differentiate between an active fire and a false detect. Smoke plumes in geostionary visible imagery can do the same.
6) How is this product created?
VIIRS channel I4 is the primary driver of the fire detection algorithmpresented in this study. The spectral response of Channel I4 (ranging from 3.55 to 3.93μm, centered at 3.74 μm) spans the wavelengths of peak spectral radiance for blackbodies emitting at temperatures between 737 and 817 K. Channel I4 is therefore well suited for distinguishing pixels containing sub-resolution combustion components from those pixels composed of cooler fire-free background areas. Channel I4 is a single gain channel with a pixel saturation temperature of 367 K. The relatively low saturation temperature, in combination with the improvedspatialresolution,canresult in frequent fire pixel saturation. This is therefore considered the most important channel characteristic influencing the development of an active fire algorithm for this data set. Another important aspect of VIIRS I4 data involves the channel's spectral placement, which is approximately 0.3μm shorter than the corresponding VIIRS M13 radiometer channel driving the baseline 750 m Active Fires ARP algorithm. This spectral offset introduces an approximately three-fold increase of the reflected solar component in the I4 channel, with potential reduction of the radiometric separation between fire-affected pixels and bright fire-free surfaces. 1
Complementing the I4 channel data, channel I5 is centered at 10.5–12.4μm and is the primary channel against which I4 is compared in order to separate active fires from their fire-free background. This is also a single gain channel with a saturation temperature of 380 K. The three remaining I-band channels (I1, I2, and I3) cover the visible (0.6–0.68μm), near-infrared (0.846–0.885μm) and the shortwave infrared (1.58–1.64μm) regions and are used in support of cloud, sun glint and water-body discrimination in the fire detection algorithm. VIIRS channels I1–I3 are only operated during the daytime portion of the orbit. 1
1Wilfrid Schroeder, Patricia Oliva, Louis Giglio, Ivan A. Csiszar. The New VIIRS 375 m Active Fire Detection Data Product: Algorithm Description and Initial Assessment. Remote Sensing of Environment. https://geog.umd.edu/sites/geog.umd.edu/files/pubs/Schroeder_et_al_2014b_RSE.pdf