Driven by a dry summer, numerous fires have been ignited in the Pacific northwest of the United States and far southwestern Canada. These fires were largely ignited by lightning from late summer storms, but continued drying in the area has both increased susceptibility to fire while also preventing rain from quenching the flames. The largest fire, Bear Gulch Fire on the Olympic Peninsula, has grown to over 9400 acres while containment holds steady at around 10%. The true color RGB from GOES-18 as displayed on CSPP Geosphere clearly shows the nearly state-wide extent of the smoke on an otherwise cloud-free day.

The GOES Fire Temperature RGB product (described here) uses a combination of the 3.9, 2.2, and 1.6 micron channels to provide a qualitative assessment of fire temperature. Since these are shortwave channels, they are subject to solar reflectance. However, the brightness temperatures of fires are so great that they can still be seen in the day. The hotter a fire is, the more emission it produces at the shorter wavelengths. When those three channels are combined into an RGB recipe, the fire hot spots change from red to yellow to white with increasing fire temperature as more influence from those shorter-wavelength channels is present.

This animation below shows the Fire Temperature RGB product in action. Note the numerous red dots across the region.

However, the Bear Gulch Fire is hardly visible via this product despite being the largest fire in the state of Washington. Located southwest of Seattle on the Olympic Peninsula, this fire is only intermittently present on the RGB product. Zooming in on the target area helps emphasize this as the bright red dot representing the fire only appears briefly over the course of a 2+ hour long animation.

The reason for this has to do, in part, with the unique terrain of the region. The GOES aerosol optical depth product vividly indicates just how extensive the smoke is across the state of Washington. However, the plume from the Bear Gulch Fire (seen as the red area in western Washington just west of Puget Sound in the below movie) has a much smaller extent than many of the other, smaller fires. This is because the mountainous terrain is largely trapping the smoke.

This trapped smoke results in a higher optical depth than in other locations, which then obscures the infrared emission from the fire. As a result, the 3.9 micron channel view of the fire is largely insensitive to the fire’s high temperatures; only occasionally does the smoke plume abate enough to enable some of that intense radiant energy to escape to space.

View only this post Read Less

1-minute Mesoscale Domain Sector GOES-18 (GOES-West) Clean Infrared Window (10.3 µm) images combined with the FDCA Fire Mask derived product (above) showed that the Dollar Lake Fire in western Wyoming produced a pyrocumulonimbus (pyroCb) cloud as it was experiencing very active fire behavior (growing in size from 1300 to 5000 acres) in the wake of a cold frontal passage on 22 August 2025. The pyroCb cloud first exhibited a cloud-top 10.3 µm infrared brightness temperature (IRBT) of -40ºC (denoted by darker blue pixels) — a necessary condition to be classified as a pyroCb — at 1908 UTC.

A non-pyroCb meteorological thunderstorm also developed just north of the Dollar Lake Fire pyroCb. While no GLM-detected lightning activity was seen with the pyroCb, several GLM Flash Points were associated with the northern thunderstorm (however, since the Flash Points are parallax-corrected, their surface locations appeared south-southwest of the parent thunderstorm).

The coldest pyroCb cloud-top IRBT was -53.49ºC at 2222 UTC, about 300 mi downwind (east) of the Dollar Lake Fire, as the leading edge of the pyroCb approached the Wyoming/South Dakota border (above). On a plot of rawinsonde data from Riverton, Wyoming (below) the air temperature of -53.49ºC occurred just above the Most Unstable (MU) air parcel’s Maximum Parcel Level (EL) of ~10.2 km.

The Dollar Lake Fire burned very hot, exhibiting a 3.9 µm infrared brightness temperature of 137.88ºC — which is the saturation temperature of the GOES-18 ABI Band 7 detectors — for about 3 hours, beginning at 1823 UTC (below).

 

The Dollar Lake Fire continues to burn today, which is very evident when you look at nearby webcams. A large smoke plume is developing and forming pyrocumulus clouds. These clouds are being picked up by our radar showing the plume continuing to spread to the east. #wywx pic.twitter.com/CzDa9L68qU

— NWS Riverton (@NWSRiverton) August 22, 2025

View only this post Read Less

At 15:28 UTC today, NOAA decommissioned the final remaining legacy POES satellite, NOAA-15. The spacecraft was launched on May 13, 1998, and entered full operations on December 15th of that same year. Amazingly, that equates to 27 years of observations from low-earth orbit; for a spacecraft with an original mission lifespan of 2 years, that is incredible! Based on available information, that should cement NOAA-15 in the record books as the longest-lived operational American polar weather satellite, a record that is unlikely to be broken anytime soon as modern constellations (such as JPSS) now adhere to more stringent deorbiting requirements.

In recent years, NOAA-15’s various instruments have become degraded and thus its data in operations has had limited utility. Additionally, its L-band HRPT direct broadcast antenna was operated at a lower power than other POES satellites, resulting in more frequent data dropouts during reception. With that said, there was still some interesting data coming from NOAA-15’s AVHRR imager as of its last active pass over SSEC’s Madison X/L-band DB antenna on 12:46 UTC on August 18th:

		[Inactive]
AVHRR Band 1: “Red Visible” [0.63 µm]	AVHRR Band 2: “Vegetation Near-IR” [0.86 µm]	AVHRR Band 3A: “Snow/Ice Near-IR” [1.61 µm]
AVHRR Band 3B: “Shortwave IR Window” [3.74 µm]	AVHRR Band 4: “Legacy IR Window” [10.8 µm]	AVHRR Band 5: “Dirty IR Window” [12.0 µm]

Notably, NOAA-15 was the final active satellite to broadcast an Automatic Picture Transmission (APT) signal. APT was an analog direct broadcast format first developed in the 1960s, and was onboard the legacy POES and precursor spacecraft. Broadcast in the VHF range around 137 MHz, APT was easy to pick up with inexpensive radio frequency equipment. It was engineered in such a way that it did not require a tracking antenna to receive, so a simple V-dipole antenna in a fixed location could pick up the signal as the satellite tracked across the sky. Given these traits, and the availability of cheap software defined radios (SDRs) since the 2010s, APT reception was a common beginner project in the satellite/radio hobbyist communities. In May 2025, I set up a temporary APT receive station in rural northern Wisconsin, using an off-the shelf V-dipole kit, SDR, filter/amplifier, and single board computer:

Raw APT was transmitted as an audio signal, one line at a time, and the final output (when decoded) contained images from 2 of the AVHRR bands. At any given time, the two bands being sent were configured by the NOAA ground station – during the daytime, bands 2 (left) and 4 (right) were the most common. The spatial resolution was downscaled to roughly 4 km, and the nature of the analog format made it susceptible to interference from many types of equipment. However, with the right hobbyist-grade equipment and a clear view between the satellite and receive station, it was possible to get fairly usable imagery via APT.

With the legacy NOAA POES constellation now turned off, NOAA’s low-earth orbiting weather satellite efforts continue with the JPSS constellation, with S-NPP, NOAA-20, and NOAA-21 all in orbit now and two additional missions planned. NOAA’s partners at EUMETSAT have been busy, with the recent launch of their first spacecraft in the MetOp Second Generation series, MetOp-SG-A1, and their prototype passive microwave sounder mini-satellite, Arctic Weather Satellite, now operational. Also, JAXA recently launched its next-generation passive microwave imager, AMSR3, on the satellite GOSAT-GW, which will replace AMSR2 on GCOM-W1.

View only this post Read Less

1-minute Mesoscale Domain Sector GOES-19 (GOES-East) Visible and Infrared images (above) showed the WNW motion of the eye of Hurricane Erin during a 7-hour period as the tropical cyclone rapidly intensified from a Category 4  storm at 0950 UTC to a Category 5 storm at 1520 UTC, north of the Leeward Islands on 16 August 2025. Erin was the earliest Category 5 hurricane on record in the Atlantic Ocean, as well as one of the Atlantic’s fastest-intensifying tropical cyclones on record (SATCON). Plots of 1-minute GOES-19 GLM Flash Points displayed abundant lightning activity within the inner eyewall of Erin’s pinhole eye. In addition, the eye exhibited a notable stadium effect — with a very small low-altitude eye seen in visible imagery, broadening to a larger high-altitude eye in infrared imagery (for example, at 1657 UTC). The coldest cloud-top infrared brightness temperatures within the eyewall were around -80 C.

A closer view of 1-minute GOES-19 Visible imagery (below) revealed the presence of low-altitude mesovortices within the eye. A pinhole eye and mesovortices within the eye are 2 satellite-observed characteristics often associated with intense Category 5 tropical cyclones. In addition, Erin was moving though an environment of weak deep-layer wind shear and traversing warm sea surface temperatures — 2 factors that favored intensification.

A GOES-19 Visible image at 1424 UTC (below) included plots of ASCAT ocean surface winds valid at that time; hurricane-force winds only extended about 10-15 miles from the center of Erin. However, significant rainfall contamination within portions of the eyewall adversely affected the quality of a few of the scatterometer winds (with 3 erroneous wind directions seen near the eye).

A Synthetic Aperture Radar (SAR) image from RCM-3 at 2229 UTC on 16 August (below) depicted a peak radial wind velocity of 123 kts in the SE quadrant of Erin (source) — this was around the time that the hurricane was undergoing an eyewall replacement cycle, and decreasing in intensity from a Category 5 at 2100 UTC to a Category 4 at 0000 UTC.

Later that evening, 1-minute GOES-19 Infrared images (below) showed that the eye of Category 4 Hurricane Erin passed about 40 miles south of Buoy 41043 — which reported hourly wind gusts of 64 kts at 0200 UTC and 0300 UTC on 17 August (the highest 10-minute wind gust was 66.1 kts at 0250 UTC).

Plots of wind speed / wind gusts / atmospheric pressure and wave height at Buoy 41043 (below) indicated that their maximum (or minimum, in terms of pressure) values occurred around 0300 UTC on 17 August (just as the eye of Erin was passing south of the Buoy).

The maximum wave height of 29 ft measured at Buoy 41043 was commensurate with the 25.32 ft significant wave height sensed by Sentinel-3A (and the 39.52 ft sensed by SWOT) farther to the south, near the northern Leeward Islands, during the morning hours (1237 UTC and 1428 UTC) on 16 August (below).

View only this post Read Less