Abstract:
In the field of climate services, characterization of rainfall extremes is useful to identify and quantify rain rates
that can trigger floods, on-farm water stagnation, excess run-off causing arable soil depletion and other natural
hazards. Delving through multiple sources of observational uncertainties, we define extreme rain events (EREs)
using the 99th percentile thresholds of daily accumulated rainfall, extracted from historical data records
(1960–2016) of manual and tipping bucket gauges. The results of the analysis show that the average amplitude
of these threshold values has been increasing in the recent years. Meanwhile, the three categories of heavy rains
exhibit an intra-seasonal timing that follows different phases of the West African monsoon. Category 1 & 2 occur
mostly in the northern Sahel, between weeks 27 and 35 of the year with an accumulated daily amount varying in
the 37–65 mm range and less than or equal to 85 mm/day respectively. In category 3, rain rates are greater than
85 mm/day, observed between 28th and 38th week-of-the-year predominantly in the southern and western of
the Sahelsub-region. For each category of ERE, high risk areas are mapped using the relative probability of
occurrence at local scale. This classification can be exploited for forecasts verification, climate model evaluations
and operational early warning services against high impact rainfall events.
