Flash floods: Observations and analysis of hydro-meteorological controls

Understanding the hydro-meteorological processes that control flash flooding is extremely important from both scientific and societal perspectives. On the one hand, flash floods rank highly among natural disasters in terms of the number of people affected and the number of fatalities. The potential for flash flood casualties and damage is increasing in many parts of the world due to the social and economic pressures on land use. Also, as the planet warms and the hydrological cycle intensifies, there is the possibility that flash floods increase (Huntington, 2006). On the other hand, the analysis of flash flood processes is important scientifically because these events often reveal aspects of hydrological behaviour that either were unexpected on the basis of weaker responses or highlight anticipated but previously unobserved behaviour (Smith et al., 1996, 2005; Delrieu et al., 2005; Archer et al., 2007; Braud et al., 2010). Flash floods are associated with short, high-intensity rainfall rates, mainly of convective origin that occur locally. Runoff rates often far exceed those of other flood types due to the rapid response of the catchments to intense rainfall, modulated by soil moisture and soil hydraulic properties. Characterising catchment response during flash flood events, thus, may provide new and valuable insights into the rate-limiting processes of extreme flood response and their dependency on catchment properties and flood severity (Carpenter et al., 2007). Moreover, local flood-producing processes may be analysed more easily in the typical small scale flash flood basins than in larger catchments where the regional combination of controls can be more important (Merz and Blöschl, 2008a,b). However, the small spatial and temporal scales of flash floods, relative to the sampling characteristics of conventional rain and discharge measurement networks, make these events particularly difficult to observe. In an investigation of 25 major flash floods that occurred in Europe in the last 20 years, only about one half of the cases were properly documented by conventional stage measurements (Marchi et al., 2010). In many cases, the rivers were either ungauged or the streamgauge structures were damaged by the event. Furthermore, even when reliable stage observations are available, extrapolation of the rating curve and changes in the cross section geometry during the flood, often render the discharge estimates highly uncertain (Di Baldassarre and Montanari, 2009). Similar considerations apply to the rainfall estimation, as the spatial and temporal scales of the events are generally much smaller than the sampling potential offered by even dense raingauge networks (Anagnostou et al., 2006). As these events are locally rare, they are also difficult to capture during classical field-based experimentation, designed to last a few months over a given region, or in