A dynamic runoff co‐efficient to improve flash flood early warning in Europe: evaluation on the 2013 central European floods in Germany. (15th August 2014)

• Record Type: Journal Article
• Authors: Raynaud, D.; Thielen, J.; Salamon, P.; Burek, P.; Anquetin, S.; Alfieri, L.
• Is Part Of: Meteorological applications. Volume 22:Number 3(2015:Sep.)
• Page Start: 410
• Page End: 418
• Publication Date: 2014-08-15
• Subjects: Meteorology -- Periodicals; Meteorological services -- Periodicals; 551.5
• DOI: 10.1002/met.1469
• Languages: English
• ISSNs: 1350-4827

Abstract

<abstract abstract-type="main" id="met1469-abs-0001"> <title>ABSTRACT</title> <p id="met1469-para-0001">Flash floods are listed among the deadliest and costliest weather‐driven hazard worldwide. Yet, only a few systems to predict flash floods run operationally in Europe. Recently, the European Precipitation Index based on Climatology (EPIC) was developed and then set up for daily flash flood early warning for an area covering most of the continent. EPIC is a purely rainfall‐driven indicator based on the prediction of statistical threshold exceedence of the upstream precipitation to provide early warning up to 5 days in advance. Its main assumption is that flash floods are directly and solely related to extreme accumulations of upstream precipitation. It does not take into account any geo‐factors such as slope and land use or processes such as initial soil moisture, which can have a significant impact on the triggering of such events. This study proposes an enhanced version of EPIC through a dynamic and distributed runoff co‐efficient which depends on the initial soil moisture. This co‐efficient, namely the European Runoff Index based on Climatology (ERIC), is used to weigh each contribution of the upstream precipitation proportionally to the initial soil moisture. The evaluation based on 1 year of daily runs proved that ERIC reaches a threat score of 0.5 if it forecasts a probability >35% of exceeding the 20 year return period of upstream runoff. This result is 0.16 higher than for EPIC. A case study of the flash flooding affecting central Europe in June 2013 demonstrated the ability of ERIC to successfully detect and locate the affected areas.</p> </abstract>