75 Tackling the Verification of Flash Floods

Monday, 8 January 2018
Exhibit Hall 3 (ACC) (Austin, Texas)
Gregory R. Herman, Colorado State Univ., Fort Collins, CO; and R. S. Schumacher

The prediction of flash floods is a notoriously challenging forecast problem, requiring not only accurate prediction of heavy rainfall magnitudes, but also on the spatiotemporal distribution of that rainfall; the hydrologic interactions between precipitation, terrain, and the land surface; and also on antecedent precipitation and its effects on soil conditions. Further exacerbating the flash flood forecast problem is the considerable difficulty in verifying flash flood events, an essential component to forecasting any phenomenon. Flash flood reports (either human or based on stream gauges) are subject to population bias, with report databases often missing transient floods in very rural areas, and also to varying reporting and report encoding practices in different regions of the United States. Flash flood warnings have similar inconsistencies associated with differing warning philosophies across weather forecast offices, different proclivities to warn rural areas, and the fact that they don’t directly correspond to observed impacts themselves. Another approach uses quantitative precipitation estimate (QPE) exceedances of Flash Flood Guidance (FFG) issued routinely by NWS River Forecast Centers (RFCs) to verify flash floods. RFCs apply different methodologies—which can often produce highly different estimates with large discontinuities across RFC boundaries—to determine their FFGs, posing a source of potential error and inconsistency as well. Finally, some approaches avoid the plethora of aforementioned political discrepancies by verifying only with respect to observed QPEs, themselves derived across the US in a consistent manner. In particular, a fixed threshold (e.g. 2”/hr.) can be used as a proxy for flash flooding, as can exceedances of thresholds defined relative to the local precipitation climatology, such as average recurrence intervals (ARIs). The use of these five approaches as metrics for flash flooding are compared, and from this, insights onto best practices for verifying flash floods will be presented.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner