Fire detection « CIMSS Satellite Blog

Magnum Fire in northern Arizona

June 13th, 2020 | Scott Bachmeier

GOES-17 “Red” Visible (0.64 µm) and Shortwave Infrared (3.9 µm) images [click to play animation | MP4]

1-minute Mesoscale Domain Sector GOES-17 (GOES-West) “Red” Visible (0.64 µm) and Shortwave Infrared (3.9 µm) images (above) showed the smoke plume and thermal anomaly (cluster of hot pixels) associated with the Magnum Fire in northern Arizona on 12 June 2020. The hottest Shortwave Infrared brightness temperatures observed were 138.7ºC (411.9 K), which is the saturation temperature for those ABI detectors. Near and immediately downwind of the fire source region, brighter-white pyrocumulus clouds were seen penetrating the top of the darker gray smoke plume. About 50-60 miles north of the fire, the smoke plume drifted over Bryce Canyon, Utah (KBYC) — but the surface visibility there remained at 10 miles, indicating that the smoke remained aloft (and automated hourly reports listed an overcast layer at 9-12 kft from 03-05 UTC).

At 2112 UTC, the Suomi NPP VIIRS Fire Radiative Power product as viewed using RealEarth (below) revealed a maximum FRP value of 142.3 MW, and a band I4 (3.74 µm) infrared brightness temperature of 367 K.

Suomi NPP VIIRS Fire Radiative Power product [click to enlarge]

On the following day (13 June), a veil of broken to overcast cirrus moved over the Magnum Fire for much of the day — but in 1-minute GOES-17 3.9 µm imagery, the fire’thermal anomaly was only completely masked for very brief periods when the clouds were at their maximum thickness (below).

GOES-17 “Red” Visible (0.64 µm) and Shortwave Infrared (3.9 µm) images [click to play animation | MP4]

Another view of the fire using 5-minute imagery from GOES-16 (GOES-East) provided quantitative products such as Fire Power, Fire Temperature and Fire Area (below) — these 3 products are components of the GOES Fire Detection and Characterization Algorithm (FDCA). These FDCA products are still being tested and evaluated using GOES-17 data before being released.

GOES-16 Shortwave Infrared (3.9 µm), Fire Power, Fire Temperature and Fire Area [click to play animation | MP4]

Posted in Fire detection, GOES-16, GOES-17, RealEarth, Suomi NPP | No Comments »

Pyrocumlonimbus cloud spawned by the Bringham Fire in Arizona

June 11th, 2020 | Scott Bachmeier

GOES-17 “Red” Visible (0.64 µm, top), Shortwave Infrared (3.9 µm, center) and “Clean” Infrared Window (10.35 µm, bottom) images, with hourly plots of surface reports [click to play animation | MP4]

1-minute Mesoscale Domain Sector GOES-17 (GOES-West) “Red” Visible (0.64 µm), Shortwave Infrared (3.9 µm) and “Clean” Infrared Window (10.35 µm) images (above) showed the formation of a pyrocumulonimbus (pyroCb) cloud that was spawned by the Bringham Fire in extreme eastern Arizona during the afternoon hours on 11 June 2020. To be classified as a pyroCb, the deep convective cloud must be generated by a large/hot fire, and eventually exhibit cloud-top 10.35 µm infrared brightness temperatures of -40ºC and colder — assuring the heterogeneous nucleation of all supercooled water droplets to form ice crystals. The pyroCb cloud then moved northeastward across far western New Mexico.

In Shortwave Infrared imagery, the fire’s thermal anomaly or “hot spot” was depicted by the cluster of red pixels — and the pyroCb cloud tops appear warmer (darker gray) than those of nearby conventional thunderstorms, due to enhanced reflection of solar radiation off the smaller ice crystals found in the pyroCb anvil (reference).

The pyroCb exhibited minimum cloud-top 10.35 µm infrared brightness temperature in the -40 to -49ºC range (shades of blue) — according to rawinsonde data from Tucson, Arizona at 00 UTC on 12 June (below), this roughly corresponded to altitudes of 10-12 km.

Plot of rawinsonde data from Tucson, Arizona [click to enlarge]

Suomi NPP VIIRS True Color RGB image, with plots of VIIRS Fire Radiative Power [click to enlarge]

A Suomi NPP VIIRS True Color Red-Green-Blue (RGB) image viewed using RealEarth (above) included plots of VIIRS Fire Radiative Power. The hazy signature of smoke drifting northward was apparent in the image. In fact, a plot of surface observation data at Springerville, Arizona (KJTC) (below) indicated that surface visibility was eventually reduced to 7 miles around 23 UTC as strong southerly winds advected smoke northward from the fire.

Plot of surface observation data at Springerville, Arizona [click to enlarge]

Posted in Air quality, Fire detection, GOES-17, RealEarth, Red-Green-Blue (RGB) images, Suomi NPP | No Comments »

When is an ABI hot (bright) spot not a fire?

May 30th, 2020 | Tim Schmit

An ABI hot (bright) spot is not a fire when it’s a fleet of solar farms. For example, recall the CIMSS Satellite Blog entry regarding solar farms in California.

ABI band 2 visible animation on May 30, 2020 (mostly) in southeastern Minnesota. Click to play mp4.

Note how some reflections are so bright that the ABI reports dark surrounding pixels. This is part of the remapping process from detector to pixel space.

A multiple-spectral ABI comparison on May 30, 2020. The rows are: band 2, band 5, band 6 band 7, band 7 – 14 brightness temp, band 14 fire mask, band 7-14 radiance difference, band 7-14 radiance difference minus the rolling average

From left to right, top to bottom the panels are:
1) ABI band 2 reflectance, dynamically scaled to enhance contrast (will appear to flicker)
2) ABI band 5 reflectance, dynamically scaled to enhance contrast (will appear to flicker)
3) ABI band 6 reflectance, dynamically scaled to enhance contrast (will appear to flicker)
4) ABI band 7 brightness temperature, dynamically scaled to enhance contrast (will appear to flicker)
5) ABI band 7 minus band 14 brightness temperature. Red indicates positive values (extra thermal energy due to the sun and fires, if present), dynamically scaled to enhance contrast (will appear to flicker)
6) ABI band 14 brightness temperature, dynamically scaled to enhance contrast (will appear to flicker)
7) ABI Fire Detection and Characterization Algorithm (FDCA, aka WFABBA) fire detection metadata mask. Fires are red, orange, magenta, and shades of blue indicating different confidence levels. Green indicates fire-free land, shades of gray indicate clouds, dark blue indicates water.
8) Radiance difference of band 7 minus band 14 radiance in band 7 space. Red indicates positive values (extra thermal energy due to the sun and fires, if present), dynamically scaled to enhance contrast (will appear to flicker)
9) Radiance difference of band 7 minus band 14 radiance in band 7 space minus a rolling average of the 5 prior frames, to highlight changes. Red indicates positive values (extra thermal energy due to the sun and fires, if present), dynamically scaled to enhance contrast (will appear to flicker).

Aside from the solar farms, water clouds show up in the difference panels due to their reflection of shortwave radiation.

H/T to Chris Schmidt for the 9-panel ABI imagery. More about quantitative ABI products, including fire detection.

The original tweet from the La Crosse WFO: “We saw some awfully bright looking “clouds” showing up via satellite in southeast Minnesota earlier this afternoon. Well after some investigation, we were able to determine they were actually solar panel arrays that the sun was hitting just right!”

Solar farms and GOES-16 ABI visible imagery from the La Crosse NWS WFO.

Posted in Fire detection, General interpretation, GOES-16, Training | No Comments »

Pier 45 Fire in San Francisco

May 23rd, 2020 | Scott Bachmeier

GOES-17 Shortwave Infrared (3.9 µm, top left), GOES-16 Shortwave Infrared (3.9 µm, top right), GOES-17 Near-Infrared “Snow/Ice” (1.61 µm, bottom left) and GOES-17 Near-Infrared “Cloud Particle Size” (2.24 µm, bottom right) [click to enlarge]

The thermal signature of a large nighttime fire at Pier 45 in San Francisco (media report) was evident in Shortwave Infrared (3.9 µm) images from GOES-17 (GOES-West) and GOES-16 (GOES-East) — the warmest 3.9 µm brightness temperature sensed by GOES-17 was 27.8ºC (at 1151 UTC), while the warmest temperature sensed by GOES-16 was only 14.2ºC (at 1146 UTC).

Note that a faint thermal signature of the fire (pixels exhibiting dim shades of white) was also apparent in GOES-17 Near-Infrared “Snow/Ice” (1.61 µm) and GOES-17 Near-Infrared “Cloud Particle Size” (2.24 µm) images. This is because those two ABI spectral bands are located close to the peak emitted radiance of very hot features such as volcanic eruptions or large fires (below).

Plots of Spectral Response Functions for ABI Bands 5, 6 and 7 [click to enlarge]

Just after sunrise, the northward meandering of smoke could be seen in GOES-17 “Red” Visible (0.64 µm) images (below).

GOES-17 “Red” Visible (0.64 µm) images [click to enlarge]

However, a larger-scale view of GOES-17 True Color Red-Green-Blue (RGB) images created using Geo2Grid (below) revealed that filaments of higher-altitude smoke were drifting southward, while the aforementioned low-latitude smoke was drifting more slowly northward.

GOES-17 True Color RGB images [click to play animation | MP4]

A profile of 12 UTC rawinsonde data from Oakland (below) explained these differences in smoke transport — winds at higher altitudes were stronger, and had a northerly component.