Rainfall risk over the city of Abidjan (Ivory Coast): First contribution to the joint analysis of daily rainfall from a historical record and a recent network of rain gauges.

In recent decades, floods are recurrent and intensifying in African major cities. Every year, rains cause material damage and human losses, in Abidjan (Ivory Coast). This study contributes to the characterization of the rain hazard in the District of Abidjan. The available data are made up of daily rainfall from a historical station "Abidjan airport" (1961-2014) and an academic network of 21 rain gauges progressively implemented in Abidjan since 2016.

 The annual rainfall observed from 2016 to 2021 on the sites of this network are representative of the interannual variability of the historical station. An analysis (date of occurrence, rainfall depth, mean intensity of wet days and number of rainy days) on the Highest Cumulative Rainfall Periods (HCRP:15, 30 and 60 days) is conducted on the long-term station, in order to characterize the periods of highest risk of flooding occurrence during the long and short rainy seasons. Date of HCRP occurrence during long rainy season is relatively stable over time. It tends to occur later in the short rainy season. In recent years, the mean intensity of wet days remains steady despite intensifying flooding.

The current network of rain gauges permits to discuss the spatial variability of the date of HCRP over Abidjan. It can be seen that the date of HCRP occurrence in the long rainy season is uniform across the area of Abidjan, whereas it appears more variable during the short rainy season. At the daily scale and for 60 days- HCRP of the two rainy seasons, the extreme quantiles are determined and their evolution analysed over the historical period.  With the current network, the experimental variograms of extreme rainfalls allows to evaluate their spatial extensions according to the rainy season.

Beyond this analysis of daily rainfall, the next step will consist in the characterisation of rainfall events at the sub-daily scale in order to update the intensity-duration-frequency curves and the design storm, which is a prerequisite for a better estimation of floods and sizing of structures.