Linked Gulf Coast and Florida Convective Storm Challenges during 16–20 May 2016

Three linked Gulf Coast and Florida convective storm forecast challenges occurred between 16–20 May 2016 during NOAA's Hazardous Weather Testbed Spring Forecast Experiment (SFE) at the Storm Prediction Center. The first forecast challenge featured an area of convection that formed over the higher terrain of eastern Mexico on the morning of 18 May, grew upscale into an organized mesoscale convective system (MCS), and crossed the international border north of Laredo, TX, near 1800 UTC 18 May. A supercell formed at the southern end of this storm complex, turned southeastward, and then moved eastward across southern Texas along a weak front boundary. As this supercell approached and crossed the Texas coast between Corpus Christi and Brownsville it died abruptly, likely because it ingested slightly drier, cooler and more stable sea breeze-modified air.

The second forecast challenge involved 286 mm of unexpected record-breaking rain that fell on Vero Beach (VRB), Florida, between 1500 UTC 17 May and 0600 UTC 18 May (all but 1 mm fell on 17 May), more than doubling the previous May daily rainfall record. The record rains in VRB occurred subsequent to the formation of a massive MCS over the central Gulf of Mexico south of Mobile, AL, between 0900–1000 UTC 17 May. This MCS, which could be linked to the earlier convection associated with an anomalously strong subtropical jet (STJ) over the Gulf of Mexico (a characteristic of an El Nino year), moved east-northeastward toward Florida.

The third forecast challenge involved a large MCS that moved eastward and grew upscale as it crossed the Texas border north of Del Rio between 1000–1200 UTC 19 May. This MCS strengthened further as it moved offshore after 1800 UTC 19 May in response to strong upper-level forcing associated with the aforementioned strong STJ. Most of the SPC SFE participants (including Bosart) expected that the upscaled MCS over the northern Gulf of Mexico would continue to move east-northeastward and would bring heavy rain due to training echoes in conjunction with strong warm-air advection along the Gulf coast as far eastward as the Florida panhandle by 1200 UTC 20 May. Wrong! Instead, the upscaled MCS transitioned into a bowing MCS that produced a 3–5 hPa cold pool with widespread surface wind gusts between 35–50 kt that resembled a low-end derecho more characteristic of the Midwest and Plains.

All three MCS events occurred in a large-scale environment that favored enhanced baroclinicity across the northern Gulf of Mexico and the southern CONUS. All three events were linked in that the evolution of antecedent convection within this favorable large-scale environment contributed to the structure and evolution of subsequent MCSs. The first severe MCS was not associated with any reported severe weather reports. Rainfall amounts with the second heavy rain-producing MCS were severely underestimated by models and forecasters alike. The third MCS produced the greatest forecaster angst because rainfall totals were forecast too high (MCS propagated too fast) and severe wind reports were much more widespread than anticipated (because of cold pool formation). This presentation will attempt to untangle what happened and why it happened.