Mesoscale extreme rainfall events in West Africa: The cases of Niamey (Niger) and the Upper Oueme Valley (Benin)

In West Africa, a sharp decrease in rainfall has occurred in conjunction with an increase in flood damage since 1970. The material damage and loss of life resulting from floods highlights the undeniable vulnerability of populations to this threat and illustrates the importance of addressing the evolution of hazardous precipitation caused by intense rainstorms. This work aims to improve our knowledge of the behaviour of extreme rainfall in West Africa by studying the sub-hourly, hourly and daily evolution of the most extreme rainfall events, a topic that is especially important to those interested in studying the links between heavy rainfall and flash flooding or inundation. This study analyses the classes of extreme rainfall events in two distinct climatic areas within West Africa using the meteorological scales relevant to rainfall processes. The study is based on two precipitation datasets recorded by dense networks of rain gauges set up within the meso-sites of Niamey (Niger, Sahelian area) and the Upper Oueme Valley (Northern Benin, Soudanian zone) from 2000 to 2010 and 1998 to 2010, respectively. The Gumbel distribution was used to analyse the frequency of the maximum rainfall series for durations varying from 5 min to 24 h. The reliability of this model was examined, and the Intensity-Density-Frequency (IDF) curves derived from it were used to estimate the critical rainfall intensities at each site. The results returned exceeded frequencies that were useful for the isolation and classification of extreme rainfall cases using temporal characteristics. The climatological results confirm the existence of a latitudinal gradient in the mean annual rainfall and number of extreme events at the mesoscale. The classification methods illustrate clear distinctions between local, meso and synoptic scale events derived from convective systems over the Sahel. In contrast, Soudanian climate conditions lead to a nesting of the phenomena involved in the formation of cloud systems, making it difficult to classify rain events in that area. However, we were able to utilize the duration of rainfall events within this zone to discriminate between types of convective systems that cause extreme rainfall. For both areas, the proportion of precipitation in an extreme event compared to total yearly precipitation served as a suitable additional criterion used to objectively identify extreme precipitation event types. (C) 2016 The Authors. Published by Elsevier B.V.