The Swan Lake Fire, located in the Kenai National Wildlife Refuge, south of Anchorage, AK initiated in June 2019 due to lightning. Over the past few months, the fire has steadily grown, and as of 20 August 2019, more than 130,000 acres have burned.
To get a close look at the fire refer to the following comparison (see imagery below) at 2248Z, 19 August 2019. Imagery compares SNPP VIIRS 3.7μm to the GOES-17 3.9μm infrared imagery. GOES 3.9μm is at 2-km while the VIIRS 3.7μm is at 375-m spatial resolution. Notice the fine details of the fire areal extent seen by VIIRS, exhibiting warm brightness temperatures (i.e. fire hotspots seen in black). In contrast, the GOES imagery does not capture the intricate fire perimeter, due to its coarser resolution and that it is affected by parallax. Parallax consists of the satellite displacement of features in the imagery, where geostationary observations over northern latitudes are far away from the GOES satellite subpoint (i.e. nadir). The geostationary imagery results in elongated pixels over the fire producing ambiguity in the fire location and perimeter.
Although GOES-17 has a coarser temporal resolution, the animation below highlights the fire at a finer temporal resolution (i.e. 1-minute data) from 2230-2330Z, on 19 August 2019. The visible (0.64μm) and infrared (3.9μm) imagery shows the evolution of the fire location, fire hotspots, and corresponding smoke. Surface observations are overlaid onto the imagery to highlight the air temperature/dewpoint, wind direction and speed, along with smoke and haze identifiers.
[Click animation link] ftp://rammftp.cira.colostate.edu/torres/JPSS_Blog_Swan_Lake_Fire_AK/g17_animation.mp4
To get an idea of the atmospheric environment aloft and near the surface, users can refer to satellite derived NUCAPS soundings that provide temperature and moisture profiles. Remote areas that lack RAOB observations can take advantage of NUCAPS soundings in the operational forecasting environment. The closest NUCAPS profile in proximity to the fire is chosen below (i.e. green dot encompassed by the white circle) at ~23Z, 19 August 2019.
The 23Z NUCAPS and 00Z, 20 August 2019 RAOB sounding from Anchorage, AK are compared below. The NUCAPS and RAOB soundings are ~45 miles apart from each other, where NUCAPS soundings provide a volumetric measurement of the atmosphere and ‘smoothes’ (i.e. averages) the temperature/dewpoint measurements within the profile. In contrast, RAOBs produce measurements along a ‘point’ throughout the atmosphere, producing a finer vertical resolution. Note the RAOB observation provides wind data (surface and aloft), while NUCAPS does not provide wind measurements.
The RAOB sounding observes a weak inversion containing a dry boundary layer and light surface winds, keeping smoke in the lower atmosphere. Precipitable water values are also low in both NUCAPS (0.41 inches) and RAOB (0.53 inches) observations, indicating a relatively dry atmosphere. The NUCAPS profile provides a general idea of what the atmosphere is like, and is sampled closest to the fire in comparison to RAOB (i.e. sounding further away from the fire). However, NUCAPS misses the low-level inversion along with the higher moisture content observed in the mid-levels, noticed by RAOB measurements.
Lastly, one cannot forget the VIIRS Near-Constant Contrast (NCC) product that provides a nighttime visible capability in support of active fires. From the SNPP VIIRS overnight pass at 1235Z, 20 August 2019, the NCC observes the emitted lights produced from the fire, along with the emitted city lights. But how can users decipher between the two features? The images below compare NCC to VIIRS 3.7um to address the question. Note the emitted lights from the fires correspond with the fire hotspots (high brightness temperatures seen in black), where the emitted city lights do not exhibit this correlation. Fire is observed in between Sterling and Cooper Landing, Alaska. Emitted city lights can be seen from Nikiski to Soldtona, AK and up north, near Anchorage, AK.