CSPPGeo software to create LightningCast probability fields is in (pre-Beta stage) testing at CIMSS (as noted in this recent blog post). In addition to creating output using GOES-R data, the software will also create imagery from Himawari-9 data. There is a RealEarth instance for LightningCast probability that includes a small Guam sector cut-out, but the CSPPGeo LightningCast software allows a cutout to be created anywhere within the Himawari-9 footprint! The image above shows LightningCast probabilities for portions of northwestern Australia and adjacent oceans/islands to the north. Himawari clean window infrared (10.4 µm) imagery for the same time (from this site) is shown below. LightningCast tells you which of the convective-looking clouds are most likely to be producing lightning in the next 60 minutes. The domain for LightningCast Probability was restricted to a subset of the Full Disk image, from -21^oS to -2^oS and from 110^oE to 135^oE.

---

Thanks to John Cintineo (CIMSS) and Levi Pfantz (CIMSS) for their work on this CSPPGeo product.

View only this post Read Less

December 2023 through February 2024 was a period of exceptional wetness over the Samoan Islands. Pago Pago received 68″ of rain during that period (this was the 4th greatest total since the 1950s!!). In the past week, however, the moisture of the South Pacific Convergence Zone has shifted away from American Samoa. The relatively clear skies that have followed mean a couple things. Pago Pago had record heat on 27 February (96^oF), and Level 2 products that require clear skies are useful again. The animation above shows MIMIC Total Precipitable Water for the 24 hours endings at 1900 UTC on 28 February. The animation below shows GOES-West Band 13 imagery overlain with Level 2 Total Precipitable Water, scaled from 1.2 to 2.7 inches. Note the trench of lower precipitable water values apparent in the MIMIC fields above has an echo in the Level 2 product. In general, convection is absent in that region but more prevalent in the higher values of total precipitable water that the trench of lower values surrounds.

The Level 2 K Index, below, has a structure very similar to the Total Precipitable Water fields above. K-index is related to the Galvez-Davison Index (as shown in this blog post). The 21-h forecast of GDI is shown at bottom. There are great similarities in the structure of the K-Index, the TPW, and the GDI.

Level 2 Products from GOES-18 can be used to approximate the Galvez-Davison Index.

View only this post Read Less

The Smokehouse Creek fire (inciweb link) as of 29 February is the largest fire in Texas history. VIIRS True-Color imagery above shows an extensive smoke plume (and a lot of other clouds!) over the eastern part of the north Texas Panhandle between 1830 and 2015 UTC on 27 February. Surface observations show very strong gusts from the west. In addition, a wind shift/cold front is moving southward into the domain by 2015 UTC, denoted by a line of cumulus cloud, with strong northerly winds behind it.

SPC’s fire weather outlook had parts of the north Texas panhandle in a critical fire weather outlook as shown below, especially because of strong wind gusts. The NWS in Lubbock has a fuel dryness image (the image for 27 February is here) and that shows the fire initiated in a region that was dry to critically dry. NWS Lubbock also has composite image for fire days (link). 27 February matches well.

The Next Generation Fire System (NGFS) has various imagery to help describe the current fire and its environment. For example, the imagery above from 2137 UTC on 27 February 2024, includes Fire Temperature RGB imagery (upper left), Day Fire RGB (upper right), GOES-West fire detection pixels (on top of True-color imagery, bottom left, and on top of a map, bottom right). This event included a large number of fire pixels detected, especially around the city of Canadian TX.

The animation of Fire Temperature RGB, below, from AWIPS, shows the evolution of the ongoing fire, and its interaction with a southward-moving cold front; when the front moves through the regions with fire (for example, near 2200 UTC in Canadian TX), the propagation of the fire switches from west-to-east to northwest-to-southeast. Satellite detection of the active fire is challenged at the end of the animation below because of increasingly thick upper-level clouds moving in from the southwest.

Night Microphysics RGB imagery, below, from the CSPP Geosphere site, show thick clouds over the fire region initially; by the end of the animation, however, hot spots in the Night Microphysics RGB (pixels that are magenta/pink) start to emerge.

---

Day Night Band imagery from early on 28 February, reveals the light emitted from the fires. The I04 shortwave infrared (3.74 µm) imagery shows the heat signatures — that can help a user unfamiliar with a location differentiate fire signatures from urban lights. Relatively clear skies at 0823 UTC (on the left, click here for a toggle between the two VIIRS images at that time) allow an unimpeded view of the fires. By 0932 UTC (on the right, click here for a toggle between the two VIIRS images at that time), however, cirrus streaming in from the southwest is interfering with the satellite’s ability to detect hot spots although the light from the fires is able to penetrate the cloud deck.

---

VIIRS imagery from during the day on 28 February 2024, below, defines the large outline of the burnscar. The scar in the visible (0.64 µm) imagery is not quite so distinct as it is in 0.87 µm and 1.61 µm. The 3.74 µm (shortwave infrared) suggests that burning is continuing on 28 February. The False Color imagery uses information from the 1.61 µm and 0.87 µm bands. The burn scar is extensive, covering almost the entirety of several Texas counties!

In a toggle between VIIRS False Color RGB imagery and the GOES-16 Land Surface Temperature (LST) derived product at 1900 UTC (below), LST values were up to 10ºF warmer (mid 80s F, darker shades of red) within darker-colored portions of the Smokehouse Creek Fire (in Roberts and Hutchinson counties) and Windy Deuce Fire (in Carson county) burn scars — compared to lighter-colored northern areas of the Smokehouse Creek burn scar (where LST values were mainly in the mid 70s F, shades of green). The presence of lingering smoke (which was incorrectly classified as Cloudy in the Clear Sky Mask derived product) prevented the derivation of LST values over all parts of those 2 burn scars.

Note that the large Smokehouse Creek Fire burn scar extended several miles eastward across the Texas/Oklahoma border, into parts of Ellis and Roger Hills counties.

---

The Smokehouse Creek Fire began during the afternoon hours on 26 February, and 1-minute Mesoscale Domain Sector GOES-18 (GOES-West) Shortwave Infrared (3.9 µm) images (above) showed the rapid eastward run during its initial 7 hours — covering a distance of about 50 miles across Hutchinson and Roberts counties in the Texas Panhandle. Rapid eastward runs of the Grape Vine Creek Fire in Gray county and the Windy Deuce Fire along the Potter/Moore county line were also seen.

1-minute GOES-18 Shortwave Infrared images along with the Fire Power, Fire Mask and Fire Temperature derived products (3 components of the GOES Fire Detection and Characterization Algorithm FDCA) provided additional quantitative information about the fire during its eastward run (below). Although the InciWeb report listed the fire origin time as 14:20 CST (2020 UTC) just north of Stinnett in Hutchinson county, the initial signature in the GOES-18 FDCA products appeared at 1856 UTC (12:56 PM CST).

The Smokehouse Creek Fire burned very hot, exhibiting a maximum 3.9 µm infrared brightness temperature of 137.88ºC (which is the saturation temperature of the GOES-18 ABI Band 7 detectors) as early as 2104 UTC in eastern Hutchinson county (below). At that time and location, the derived Fire Power value was 3678.88 MW and the derived Fire Temperature value was 1742.32 K.

A longer (53-hour) animation of 5-minute GOES-16 (GOES-East) Shortwave Infrared images — centered on the Smokehouse Creek Fire — from 26-28 February (below) illustrated (1) the initial eastward run of the fire on 26 February, (2) the rapid increase in areal coverage and eastward expansion into far western Oklahoma on 27 February, as westerly winds increased in speed during the day (3) the abrupt southward expansion following the passage of a strong cold front late in the day on 27 February — note the shift from westerly to northerly winds in a plot of surface data from Borger Hutchinson County Airport (KBGD), and (4) the periodic masking of the fire signature as a patches of mid/high clouds moved across the region on 28 February.

View only this post Read Less

CIMSS Scientists who work with JPSS data had numerous presentations at the American Meteorological Society’s Annual Meeting held at the end of January in Baltimore. This blog post discusses a poster by Tom Greenwald who (along with co-authors) investigated how microwave data from AMSR-2 can be used to estimate sea ice concentration in regions where very high-resolution Synthetic Aperture Radar (SAR) data are not present. The AMSR-2 images were validated using very high-resolution Landsat visible imagery; that is, this was done in regions of clear skies to show that the microwave data would be producing useful information in regions of clouds. The high-resolution AMSR-2 Sea Ice Concentration data are used, as the poster notes, to “track the ice edge and marginal ice zone (MIZ) with a degree of confidence not achieved with the standard AMSR2 sea ice concentration.”

A dedicated CIMSS Satellite Blog reader might recall past blog posts that also discussed this technique, here and here. AMSR-2 gives high-resolution, quality sea ice detection to augment SAR observations. Future work includes a more thorough comparison between this product and legacy products. A Quick Guide on this product is available here.

View only this post Read Less