Continuing this blog’s recent theme of discussing recent events in West Texas, a tornadic supercell has formed near Muleshoe, Texas, in the panhandle region near the Texas / New Mexico border. Both CONUS and mesoscale sector scans from GOES-19 well-captured the initiation and further development of this storm.

The late morning surface analysis (1500 UTC, from NOAA WPC) shows the potential for this event. Multiple boundaries are present in the high plains of Texas and New Mexico: a southward-sagging cold front and a series of westward propagating outflow boundaries from existing central plains weather systems. These outflow boundaries are shown as dashed brown lines in the surface map.

Any of these boundaries is capable of initiating convection as they propagate along their direction of travel as both cold fronts and outflow boundaries represent the leading edge of cool, dense air that can undercut and lift warmer air. It is worth noting that the panhandle air is also somewhat moister than is often seen, with dew point temperatures in the low 60s. A very strong moisture gradient can also be seen along the dryline: witness those southern New Mexico dew points in the single digits or even below zero!

The atmosphere was primed for convection, as can be seen in a NUCAPS sounding from the early afternoon for the Texas panhandle. The best-estimate CAPE from the combined microwave and infrared sounder retrieval was 1313 J/kg at that time, which compares well with the 1900 UTC special radiosonde launched from Amarillo, which measured a mixed-layer cape of 1476 J/kg.

One might expect that the convection would be focused along the dryline when it is that strong. However, the GOES-19 CONUS true color satellite imagery shows that convection is focused along the cold front and outflow boundaries.

This is a somewhat unusual occurrence, as convection in this part of the continental United States is frequently induced from boundaries propagating eastward. Note that the most significant cell forms where a north-south outflow boundary intersects the east-west cold front. The constructive interference induced from these colliding boundaries can often induces stronger lifting, which is clearly evident here.

NWS and NOAA personnel coordinated to ensure that a mesoscale sector was available for this region, providing minute-by-minute snapshots of the region. The CIMSS ProbSevere product was also able to identify regions of concern for public safety and property. At 2102 UTC, a large multi-vortex tornado was reported from this cell (source: NOAA Storm Prediction Center)

By late afternoon, numerous other cells had initiated. Many of them carried the classic indicators of strong convection such as overshooting tops and enhanced V structures that can be seen on the Band 13 imagery. The largest cell in the center of the animation below was the original cell. Given the somewhat unusual direction that the initiating boundaries were propagating, it might not be too surprising that this cell is moving to the south rather than a more typical east-northeast direction as is shown in the 1 minute mesoscale imagery.

A closer look at the evolution of this cell also reveals something interesting (and somewhat unusual). With one of the factors forcing convective initiation today being the eastward movement of pre-existing boundary layers, a strong updraft formed beneath the anvil of the initial cell along one of those boundaries, forming its own overshooting top that penetrated through the anvil. It then moved westward and merged into and strengthened the existing updraft.

View only this post Read Less

SPC Storm reports for 1200 UTC 22 April – 1200 UTC 23 April 2025 (here) included an observation of very strong winds at the Midland Airport as shown below. (See also this link).

What did the satellite imagery for this storm show? GOES-16 CONUS visible imagery, below, shows the evolution of the thunderstorm responsible for the downburst. As noted in the storm reports, it’s weakening.

Careful inspection of the visible imagery at the end of the animation suggests the arc of an outflow boundary to the west of Odessa. Perhaps more than one strong downdraft was occurring in this environment.

GOES-19 Infrared imagery (Band 13, 10.3 µm) also shows the weakening convection.

GOES-West (GOES-18) also viewed the convection. However, west Texas is outside the GOES-West CONUS/PACUS sector with its routine 5-minute scanning, and only Full Disk imagery (every 10 minutes) is routinely available.

What was the thermodynamic structure over west Texas at this time? Midland did launch a radiosonde, and the 0000 UTC 23 April 2025 sounding is shown below, and also available here, from the handy Wyoming Sounding website. Of note is the very steep lapse rate from 500-mb down to the surface!

Both NOAA-21 (at 1925 UTC) and NOAA-20 (at 2014 UTC) overflew west Texas late in the afternoon on 22 April 2025, and NUCAPS profiles were produced over this unstable environment. The slider below compares the lapse rate (^oC) between 800 and 600 mb (imagery was created with polar2grid as described in this blog post); the atmosphere shows very strong lapse rates over west Texas, in agreement with the one radisonde above!

---

NUCAPS Sounding Availability fields from AWIPS on 22 April 2025 (below) show the following points available. Over west Texas, valid profiles (in green: that is, the retrieval converged to a solution) are indicated.

NUCAPS profiles at the closest point northwest of Midland are shown below at 1927 UTC (NOAA-21) and 2016 UTC (NOAA-20). As in the KMAF radiosonde above, a neutral atmosphere up to 500 mb is apparent at both times.

Scott Bachmeier/CIMSS notes that the GOES-19 Mesoscale Sector did a great job of capturing the small outflow boundary highlighted above. The animation below is from the CSPP Geosphere site.

---

Thanks to Lee Cronce, CIMSS, for his help in getting the (now historical) NUCAPS profiles to plot.

View only this post Read Less

1-minute Mesoscale Domain Sector GOES-19 (GOES-East) Visible images with an overlay of the Fire Detection and Characterization Algorithm (FDCA) Fire Mask derived product (above) displayed the rapid growth and movement of the Jones Road Wildfire in eastern New Jersey on 22 April 2025. The initial GOES-19 FDCA fire detection occurred at 1616 UTC. In addition, the maximum 3.9 µm infrared brightness temperature reached 137.77ºC — the saturation temperature of GOES-19 ABI Band 7 detectors — for several minutes beginning at 2109 UTC. Just north of the wildfire, westerly winds at Toms River (KMJX) occasionally gusted to 18 kts (21 mph) in the wake of a cold frontal passage, which likely played a role in the rapid eastward run of the fire toward the Garden State Parkway (that was subsequently closed to traffic for most of the following overnight hours, as the fire jumped the Parkway). Note that the dense cirrus clouds drifting over the area occasionally attenuated the fire signature enough to prevent FDCA Fire Mask detections.

In contrast, the more sensitive GOES-19 Next Generation Fire System (NGFS) (below) consistently displayed a fire signature every minute during the same 7-hour period as the FDCA shown above (in spite of the dense cirrus clouds moving overhead). The initial NGFS fire detection occurred at 1617 UTC. In addition to closure of a portion of the Garden State Parkway and a few other roads, evacuation orders were issued for approximately 5000 residents.

=====

The fire continued burning overnight, followed by an increase in intensity during the morning/afternoon hours on 23 April (below). As a sea breeze began to advance inland, the transition to southeast winds transported some of the wildfire smoke north/northwestward — and the visibility dropped as low as 2.5 miles at Toms River KMJX and Lakehurst Naval Air Station KNEL. As a result of this smoke, Air Quality Alerts were issued for parts of eastern New Jersey.

A 30-meter resolution Landsat-9 “Natural Color” RGB image from the RealEarth site (below) revealed the areal coverage of the Jones Road Wildfire burn scar (darker shades of brown), along with a few areas of ongoing fire activity (brighter shades of pink/red).

=====

In a 2-day animation of GOES-19 Fire Temperature RGB images created using Geo2Grid (below), the thermal signature (shades of red to bright yellow) was very pronounced during the day on 22 April (2116 UTC image) as wildfire rapidly grew and spread eastward — and again during the first half of the day on 23 April (1421 UTC image), before becoming more subdued later in the day as fire suppression efforts increased containment. The thermal signature was relatively subtle during the day on 24 April, except for one brief flare-up in the early afternoon (1641 UTC image).

View only this post Read Less

The Madison DBPS site (and other DBPS sites as detailed here) allows a user to view — in near-real time — derived MIRS imagery at different frequencies, such as the 165 GHz imagery below from NOAA-20 at 0728 UTC on 21 April 2025.

You’ll notice that (at present), a colorbar is not embedded within the image. A user who wants a colorbar, however, can easily access the data and create an image that includes brightness temperatures with a colorbar by using polar2grid — the software that is being used at the Direct Broadcast site to create the imagery above. (And yes, there are plans to embed a colorbar in the imagery at the direct broadcast site). How do you do this yourself?

You can access the data files used to create the imagery at two different locations, either at the direct broadcast site at CIMSS, i.e., the week-long storage at https://bin.ssec.wisc.edu/pub/eosdb/j01/ ; the datafiles you need are under atms, then a year/day/time, and Environmental Data Records, in this case: https://bin.ssec.wisc.edu/pub/eosdb/j01/atms/2025_04_21_111_0726/edr/ ; the files in that directory are shown below. You want the IMG file. This file contains all the data in the swath viewed by the direct broadcast antenna.

NPR-MIRS-IMG_v11r9_n20_s202504210728030_e202504210742266_c202504210807170.nc	 
NPR-MIRS-SND_v11r9_n20_s202504210728030_e202504210742266_c202504210807170.nc	

You can also find similar files online at (for example) the Amazon Webservices NOAA-20 bit bucket ( https://noaa-nesdis-n20-pds.s3.amazonaws.com/index.html ). There, you can find a collection of NPR-MIRS-IMG files on 21 April 2025 ( https://noaa-nesdis-n20-pds.s3.amazonaws.com/index.html#NPR_MIRS_IMG/2025/04/21/ ); the MIRS data in these files is from smaller granules than at the Direct Broadcast site. The map below that shows NOAA-20 orbits on 21 April 2025 (from this site), means data from 0730 to 0737 UTC (approximately) is needed.

The screencapture below shows some of the files needed (of course, a simpler way to do this is to write a Python script that goes and fetches the files!). The first file on this page has data from 07:33:23.0 to 07:33:54.6 and the last file shows data from 07:38:43.0 to 07:39:14.6.

I downloaded some of these (plus others on the previous page that were from before 07:33) to my computer. The listing of the files is shown below. Now create imagery with polar2grid (free software that is downloadable here; a registration may be required. The information gathered via that registration is used only to communicate changes to users).

---

The first thing to do was to create a grid that approximated the one shown at the top of this blog post, using the polar2grid command p2g_grid_helper.sh. The downloaded polar2grid software went into this directory (~/Polar2Grid/polar2grid_v_3_1/) and the commands run below are being done in the bin directory just underneath the main polar2grid directory (i.e., ~/Polar2Grid/polar2grid_v_3_1/bin).

./p2g_grid_helper.sh MADISON -85.0 43.0 4000.0 -4000.0 980 760 > Madison.yaml

The command above creates a grid centered at 43.0^oN, 85.0^oW, with a grid spacing of 4 km (4000 m) in the E-W and N-S directions, with a size of 980×760 pixels. The script output is stored in a yaml file (that polar2grid will interpret to re-grid data).

/polar2grid.sh -r mirs -w geotiff --list-products-all -f /path/to/downloaded/data/NPR-MIRS-IMG_v11r9_n20_s2025042107*.nc

The polar2grid command above lists the products that can be created. Polar2grid recognizes the data input as MIRS data — if you tell it that is the input by specifying the mirs reader : -r mirs. In addition, you can run a query about what kind of imagery can be created as shown below.

Part of that lengthy output stream from the command above is shown here. For this post I want to create 165h GHz imagery, and we’ll also create 31v and 88v GHz brightness temperatures. The polar2grid command to do that is below: the -p flag informs the software to create a list of products: brightness temperatures at 165 GHz (horizontally polarized observations), and at 88 and 31 GHz (vertically polarized observations). The data are regridded onto the ‘MADISON’ grid created above; recall that output from that grid-creation software was stored in the Madison.yaml file. All data files starting at 0700 (in this case, 072659, as listed above) will be appended by the polar2grid software). The second command adds a colormap to the created tif files.

./polar2grid.sh -r mirs -w geotiff -p btemp_165h btemp_88v btemp_31v -g MADISON --grid-configs ./Madison.yaml -f /path/to/downloaded/data/NPR-MIRS-IMG_v11r9_n20_s2025042107* 

./add_colormap.sh ../colormaps/amsr2_89h.cmap noaa20_atms_btemp_*_20250421_072659_MADISON.tif

The very flexible add_coastlines shell script will (1) draw coastlines, (2) draw a latitude/longitude grid and (3) insert a colorbar. The invocation of the add_coastlines.sh is below for all three .tif files created.

./add_coastlines.sh --add-coastlines --add-borders --add-colorbar --colorbar-text-size 14 --colorbar-height 48 --add-grid --grid-D 10.0 10.0 --grid-d 10.0 10.0 --grid-text-size 12 noaa20_atms_btemp_165h_20250421_072659_MADISON.tif
./add_coastlines.sh --add-coastlines --add-borders --add-colorbar --colorbar-text-size 14 --colorbar-height 48 --add-grid --grid-D 10.0 10.0 --grid-d 10.0 10.0 --grid-text-size 12 noaa20_atms_btemp_88v_20250421_072659_MADISON.tif
./add_coastlines.sh --add-coastlines --add-borders --add-colorbar --colorbar-text-size 14 --colorbar-height 48 --add-grid --grid-D 10.0 10.0 --grid-d 10.0 10.0 --grid-text-size 12 noaa20_atms_btemp_31v_20250421_072659_MADISON.tif

The add_coastlines.sh output is a png file. The 165h GHz brightness temperature that matches the one up top (albeit on a different grid) is shown below, but now it has a colorbar!

The animation below compares the three images created, 31, 88 and 165 GHz. As expected, resolution increases as frequencies increase (because larger frequencies mean more energy). Note also how cold the 31 GHz Brightness Temperatures are over open ocean compared to 165 GHz! A conclusion can be that ocean emissivity is a function of frequency!

Note also that colorbar values are different above for 31 GHz and 88/165 GHz. These values are controlled within polar2grid in this file: polar2grid_v_3_1/etc/polar2grid/enhancements/generic.yaml ; within the file you will find (changeable!) limits for 31v, 88v, and 165h as shown below. By default, 31v is defined as 140-300, as you see above, and 88v and 165h are both defined as 150-300.

A similar method to control how CrIS data are displayed in polar2grid is in this blog post.

View only this post Read Less