Strong Tornadoes Impact Great Plains

By Tim Wagner • 24 April 2026

In the afternoon and evening of 23 April 2026, numerous tornadoes touched down across the Great Plains of the continental United States. The storm reports gathered by NOAA’s Storm Prediction Center show a line of tornadoes extending from central Iowa down to north central Oklahoma, accompanied by numerous reports of large hail and damaging winds. While full damage surveys are pending, it appears that some of the most destructive storms were on the south end of the line. The Oklahoman newspaper in Oklahoma City has video of the tornadoes, including one that damaged structures on Vance Air Force Base near Enid in the north central part of the state. Fortunately, as of press time, no fatalities have been reported.

The storms were initiated by a particularly strong dry line. In fact, we’ll begin by highlighting just how intense this dry line was: it was so strong you could see it from space! The dry line was clearly visible on the GOES-19 (GOES West) Band 13 Infrared Window Channel, which isn’t a channel where you’d expect to see a dry line. Under clear sky conditions, you would think that two locations with the same surface temperature would have the same brightness temperature. However, look at the following image from 2100 UTC (4:00 PM CDT). There is a definite boundary running along from north-northeast to south-southwest. However, temperatures on either side of this boundary are a consistent 86 F. It turns out that there is so much water vapor in the east that it’s causing the brightness temperature in the east to be a little colder than in the bone-dry west, causing the right side of the image to be a little lighter than the left side.

Dry lines serve as a useful trigger for convective initiation as the dry air is denser than humid air. Compare a pair of special radiosonde launches from 1800 UTC (1:00 PM CDT) from Amarillo, TX (in the heart of the Texas Panhandle) and Lamont, OK (in the north central part of the state) as depicted on the SPC’s Observed Sounding Archive. This is a slider, so you can drag the center line back and forth to compare the differences between the two. (It may be challending to see the light gray slider handle in the middle of the image, but it’s there; drag it back and forth to see how conditions change and to ensure that you’re not seeing the Lamont hodograph with the Amarillo skew-T). Both show nearly identical profiles between 850 mb and the tropopause but they have very different low level structure. In Amarillo, on the dry side of the dry line, the surface dew point depression is a whopping 62 F! That results in clear skies and very little chance for convective development. But in Lamont, by contrast, convection is much more likely. That profile is a classic “loaded gun” sounding, with a low-level capping inversion separating a modestly moist planetary boundary layer from dry air aloft. If an external trigger can hoist those moist parcels above the cap, then they’ll accelerate upward on their own. Note also that both sites have clockwise curved hodographs, indicating wind shear that is favorable for supercell development.

Let’s take a look at the cell initiation. To start with, we’re going to being a couple of hours before the IR satellite image above to gauge what the preconvective environment was like. Seen below is Band 2 (visible red) from GOES-19. Remember, this is the band with the highest spatial resolution of any of the bands on the GOES Advanced Baseline Imager (ABI). Here we see a single frame at 1900 UTC (2:00 PM CDT). The northern part of this view is covered by so-called fair weather cumulus. There are a couple of indicators that things may be preparing to change. The first is the clear gradient in the dew point. Just left of center of the image below is Enid, OK, which at this time is reporting 84/66. But just 50 miles to the northwest is Alva, OK, which is reporting 91/43. That’s a drop in the dew point of nearly 1 F for every 2 miles. The winds are generally southerly in the south and east parts of the image, but they’re more westerly in the west and north parts. This is intensifying the moisture gradient as the southerly winds are advecting moist Gulf air while the westerly winds are bringing in air from the arid high plains. (In fact, you can see this play out in the advected layer precipitable water product provided by our friends at CIRA).

As we advance forward two hours, our guess about the intensification of the water vapor gradient was correct. Enid got a little moister at 87/65, but Alva got a whole lot drier and is now 93/28. That’s now a 37 degree change in the dew point over just a 50 mile difference. We’re also starting to see a little more convergence along the dry line as the Alva winds have veered slightly while the Enid winds have backed just a bit. This is starting to focus convection along the moisture boundary which can be seen as a line of cumuli stretching northeast to southwest across the image.

If some of these parcels can penetrate high enough, we’re going to see some explosive development. One of the GOES-19 mesoscale sectors was trained on this area, so we have a minute-by-minute view of the evolution of the environment. This loop of Band 2 runs from 2000 UTC (3 PM UTC) all the way to 0100 UTC (8 PM UTC) giving us the highest possible spatial and temporal resolution view of the initation and development of this event. The long shadows of the setting sun help make the overshooting tops readily apparent.

The cell that produced the Enid tornado formed at the very south of the line just as the sun was setting. Here’s the same time period as the above loop, but this time as seen by Band 13, the infrared window. Here, the overshooting tops are clearly evident as cold (red) spots against the slightly warmer (yellow/orange) anvils. We also see evidence of enhanced v signatures as the upper level winds divert around the overshooting tops.

One of the easiest ways to identify where the initiation boundary is via satellite imagery is the Air Mass RGB product. In the loop seen below, the warm, moist air mass that originated over the Gulf is green while the dry air of the high plains is purple. In this loop, which focused on the 1900 to 2300 UTC (2:00 PM to 6:00 PM CDT) run, you can see the boundary tighten and intensify before the cells even initiate. This is the area to target which monitoring where initiation will take place.

The Day Convection RGB can also be used to monitor where cells are initiating and undergoing rapid development. Areas of yellow highlight regions where active convective development is taking place. However, users have to be careful as this is a daytime-only product and this event straddles the transition from day to night. The following loop shows the issues with this as it stretches from 2130 UTC to 0100 UTC (4:30 PM to 8:00 PM CDT), and the clouds generally all become the same shade of magenta as the sun sets. The GOES-19 Geostationary Lightning Mapper (GLM) five minute flash density is overlaid on the top of this product to highlight where lightning is taking place.

____________________

1-minute GOES-19 Infrared Window (10.3 µm) images, with/without overlays of 1-minute GLM Flash Points (white dots) and METAR surface reports (cyan), from 2109 UTC on 23 April to 0208 UTC on 24 April (courtesy Scott Bachmeier, CIMSS) [click to play MP4 animation]

1-minute GOES-19 Infrared images that included an overlay of 1-minute GLM Flash Points (above) provided a closer view of the lightning activity associated with the severe thunderstorms in southern Kansas and northern Oklahoma. Also included at the end of the animation are plots of METAR surface reports; the two METAR sites in the vicinity of Enid OK are Vance Air Force Base (KEND) just south-southwest of the city center, and Enid Woodring Regional Airport (KWDG) just east-southeast of the city center (map). Of note was the Peak Wind of 93 kts (107 mph) at KEND, which occurred at 0111 UTC (this was also the start time of the EF4-rated Enid tornado that was on the ground for about 37 minutes). The coldest cloud-top infrared brightness temperature during the Enid tornado was -79.15ºC at 0138 UTC — which represented a ~2.0 km overshoot of the Most Unstable (MU) air parcel’s Equilibrium Level (EL), according to a plot of 0000 UTC rawinsonde data from Norman OK.

1-minute GOES-19 Infrared images with time-matched plots of SPC Storm Reports (below) included a report of 4.00 inch diameter hail in southern Kansas, 3.75 inch diameter hail in far northern Oklahoma, and the tornado with a 107 mph wind gust near Enid.

1-minute GOES-19 Infrared Window (10.3 µm) images with time-matched (+/- 3 minutes) SPC Storm Reports plotted in white, from 2145 UTC on 23 April to 0135 UTC on 24 April (courtesy Scott Bachmeier, CIMSS) [click to play animated GIF]

____________________

Of course, the CIMSS ProbSevere product was tracking this event as it evolved. Here’s a snapshot at 0115 UTC (8:15 PM CDT) depicting how ProbSevere assessed the storm that produced the Enid tornado. Note the high confidence that ProbSevere has in the presence of severe activity: overall severe probability is 92% while tornadic probability is 65%.

Today, Friday 24 April, is another severe weather day in Oklahoma, so it may be a few days before the damage survey crews are able to assess the number and severity of these tornadoes. Regardless, it was an intense day that was well-captured by a variety of satellite products.

Categories: General interpretation, GLM, GOES-19, Lightning, ProbSevere, Red-Green-Blue (RGB) images, Severe convection