Diagnosing the Meteorology of 'Double Impact' Tornado/Flash Flood Events

While both tornadoes and flash floods individually present public hazards, when the two threats are both concurrent and collocated (referred to here as TORFF events, short for “tornado and flash flood”), a unique set of concerns arise that can further jeopardize public safety. Among these unique concerns for dual threat scenarios is a conflict between the recommended lifesaving actions for each individual hazard, which can increase public confusion and lead to sub-optimal precautionary responses. This can be further exacerbated when one threat is not identified in a timely manner, or is over- or underemphasized relative to the other. It is therefore critical that a forecaster is able to identify dual threat scenarios such as TORFF, and effectively communicate the pertinent risks that the phenomena present to public safety. The ability to accurately identify TORFF scenarios is essential to this task, and doing so requires an improved understanding of the distinct meteorological conditions that give rise to TORFF events as opposed to single-hazard scenarios. Among tornadoes and flash floods, the signal for tornadogenesis is often the more apparent of the two at the mesoscale and synoptic scales; an important first step in TORFF forecasting is thus distinguishing TORFF scenarios from those that produce only tornadoes (termed TOR events).

Two complementary methods were employed to further explore the meteorological conditions that distinguish TORFF events from their TOR counterparts. Tornado tracks from the Storm Prediction Center's Severe Geographic Information System and flash flood reports U.S. Flash Flood Observation Database were used to identify TOR and TORFF events over the contiguous United States during the 2008-2013 period. The first method, full field analysis (FFA), used the North American Regional Reanalysis (NARR) to create event-centered composites for both the TOR and TORFF event classes. This method is of particular utility in assessing tornado threat, as severe weather potential, including tornadoes, are known to depend on satisfying absolute environmental thresholds. Atmospheric fields of choice in this method focused on 850 hPa winds, temperature, and temperature advection; precipitable water (PWAT); most unstable convective available potential energy (MUCAPE); and bulk shear from 0-6 km. The second method also used event-centered compositing, but instead used the National Oceanic and Atmospheric Administration's Second Generation Global Ensemble Forecast System Reforecast (GEFS/R) data to create local standardized anomalies (LSAs) to assess local departures from climatology for numerous atmospheric fields for both the TOR and TORFF event classes. The combination of reanalysis and reforecast data between the methods was chosen to maximize research information and utility for the forecaster; the reanalysis assists with understanding of the true conditions, while the reforecast data more closely represents data that a forecaster would have at their disposal for TOR and TORFF forecasting. This latter method is also considered to be of particular utility in assessing the flash flood threat, since flash flooding depends not only on absolute rainfall received, but local hydrology, surface characteristics, antecedent rainfall, and other characteristics that are more acclimated to the local climatological conditions. The LSA method examined many different fields, including horizontal winds and moisture from the surface to 500 hPa; shallow and deep layer shear; vertical motion at 850 hPa; surface temperature and pressure, model soil moisture; and PWAT, CAPE, and 850 hPa temperature advection as explored in FFA.

The results of both methods largely agreed on some key findings. TORFF events were found to occur in moist, and anomalously moist environments; both methods indicated that the PWAT near TORFF events was, on average, both higher and more anomalously high relative to climatology than TOR events. TORFF events were also seen to have more 850 hPa warm air advection than TOR events, both in absolute and climatological anomaly frameworks; a similar, higher CAPE in TORFF environments finding was observed in both cases as well. 850 hPa vertical motion anomalies were perhaps the strongest signal in the LSA analysis, with significant synoptic scale forcing for ascent in both TOR and TORFF cases, but much stronger large-scale forcing in the TORFF events. LSAs also identified TORFF events as existing in anomalous southerly winds- more anomalous than TOR cases- at all levels examined, in addition to shallow and deep layer meridional wind shear. TORFF events were also found to occur in more anomalously easterly near-surface zonal flow. Overall, results indicate stronger synoptic scale forcing in TORFF cases compared with TOR scenarios, as regardless of the atmospheric field examined, LSAs indicated, and FFA corroborated, departures from climatology were larger in TORFF situations.