Is the current flood of data enough? A treatise on research needs for the improvement of flood modelling

Abstract

Flood disasters account for more than 40% of all natural hazardsworldwide and about half of all deaths caused by natural catastrophes(Ohl and Tapsell, 2000). For instance, in 2010, according to theEmergency Events Database (EM-DAT, 2011) floods were responsible forthe loss of more than 8000 human lives and affected about 180 millionpeople. The catastrophic floods in Pakistan and Australia are the mostrecent examples of a worldwide increasing human exposure to flood risk.Figure 1 reports the number of fatalities caused by floods in Africa inthe period 1950–2009 (EM-DAT, 2011).The diagram shows that the number of flood fatalities in Africa hasdramatically increased by more than an order of magnitude in the last60 years (Figure 1). These dramatic figures indicate a need for urgentactions to decrease these staggering numbers (Di Baldassarreet al.,2010a).The flood risk is likely to grow further because of many factors,such as changing demographics, land use changes, climate variabilityand change, technological and socio-economic conditions, industrialdevelopment, urban expansion and infrastructure construction in flood-prone areas as well as unplanned human settlement in floodplains(UN-ISDR Scientific and Technical Committee, 2009). To mitigate thecontinuously increasing flood risk the currently proposed approach isintegrated flood management (more towards ‘living with floods’) that hasreplaced the more traditional flood defense approach (‘fighting floods’).This approach aims to minimize the human, economic and ecologicallosses from extreme floods while at the same time maximising the social,economic and ecological benefits of ordinary floods (e.g. UNESCO-IFI:International Flood Initiative).In this context, the aim of this commentary is twofold. First, to criti-cally discuss the current opportunities for improving flood modelling,which are offered by the increased availability of new data sources,such as the utilisation of remote sensing data of higher resolution andthe increasing availability of globally and freely accessible space-bornedata. Second, to discuss knowledge gaps and research needs to fullyand properly exploit this current ‘flood of data’ (Lincoln, 2007) andimprove flood modelling. This leads to the proposed elements of a futureresearch agenda to support the scientific basis of integrated flood management.