Stereoscopic views of severe convection over Nebraska

By Jochen Kerkmann (EUMETSAT), Scott Lindstrom (CIMSS), Dan Lindsey (CIRA)

On 8 June 2020, strong convection developed over central Nebraska, a very rural area, late in the afternoon. Storm Prediction Center (SPC) storm reports included several tornadoes, large hail up to 10.6cm (4.25in), and strong wind gusts. The first hail (1.5 inches) was reported at 23:32 UTC, the first tornado was observed just 15 minutes later, at 23:47 UTC.

The storms developed ahead of a pronounced upper-level trough over the Rockies, in an environment of strong shear and moderate upper-level south-westerly flow (see Figure 1 and 00:00 UTC radiosounding from North Platte (Nebraska). Credit: University of Wyoming).

Clouds from Tropical Storm Cristobal can be seen over the south (Missouri, Arkansas); clouds from the upper-level trough are seen over Wyoming (Figure 2). In addition, a pronounced dryline (north-south oriented) separated dry air in the west from moist air (TPW values of 40mm and more) in the east. Convective initiation occurred over central Nebraska, in an area of marked humidity (darker blue area in the 24h-Microphysics RGB).

An overview of the explosive convective outbreak over Nebraska is given in Figure 3, which shows the animation of the 'clean' window band IR10.3. Numerous convective storms can be identified, some of them moving upstream against the upper-level flow (so-called backbuilding storms).

Mesoscale domains from both GOES-16 and GOES-17, viewed the developing convection, enabling fine spatial and temporal-scale stereo viewing of the convection. Search for past Mesoscale domain sectors or see locations in the past year.

A two-panel comparison of GOES-17 and GOES-16 visible images (VIS0.6) during the period 23:11-00:48 UTC is shown below (Figure 4).

The images are displayed in the native projection of each satellite. While GOES-16 gives a better view (more contrast) of upper level features like overshooting tops and above-anvil cirrus plumes (AACP), GOES-17 sees lower level features and cloud development under the anvil, i.e. the western flanks of the storms. Having both views, one could create a 3D animation using, for example, the so-called anaglyph technique (which requires viewing glasses, not shown here).

Figure 5 shows a side-by-side view of the colour-enhanced IR10.3 band and the Convection RGB, for the first 1.5 hours of the storm development (time given in the upper right corner).

Three major storms can be seen developing on a line (probably a convergence line, see Figure 6). Each storm exhibits distinct overshooting tops with top temperatures below -70°C and very narrow cold U signatures (not very visible).

The Convection RGB shows a number of yellow features (see Figure 7), which could indicate small ice particles. However, when comparing the Convection RGB to the IR10.3 image the yellow dots strongly correlate with the coldest parts of the CBs. Thus, in this case, the yellow colour is mainly due to very cold tops and not to small ice particles.

To confirm the absence of small ice, we looked at the IR3.9 animation (Figure 8), where the storms appeared in white to light grey colours, which indicates low IR3.9 reflectivities due to large ice particles.

In summary, from the VIS0.6, IR3.9, IR10.3 and Convection RGB loops the following timeline of satellite signals can be drawn: -

1. First small ice N/A (no small ice observed)
2. First very cold overshooting top at 23.41 UTC
3. First plume (AACP) at 23:41 UTC
4. First weak cold U at 23.51 UTC

Given that the first clear signal of overshooting tops was at 23:41 UTC, nine minutes after the first hail report, there was no warning lead time. The extremely rapid development from convective initiation to the first occurrence of hail/tornado made this case very difficult to nowcast.

Additional content

Stereoscopic views of severe convection over Nebraska (CIMSS Blog)