Abstract Using half-hourly rainfall data from 14 automated weather stations over central Africa, rainfall events [≥0.1 mm (30 min)−1] characteristics are explored. A total of 10 096 wet events (WEs) were identified and classified into six storm types (STs), mostly discretized by their duration and intensity. ST 1 is very short (<1.5 h) with low rainfall intensities over a small area and contributes the least to the total rainfall (7%) but is by far the most frequent (70% of the WEs). ST 2 is short (∼1.5 h) and sudden with very intense rains and of medium spatial scale (<200 km). ST 3 is short and of medium scale too but with low intensities and rainfall amounts. ST 4 and ST 5 are of large scale and long (∼3–4 h), with high rainfall amounts so that they contribute the most to the total rainfall (29% and 20%). Last, ST 6 is the largest, longest, and rainiest, although moderately intense. A complementary classification is performed on lagged gridded rainfall fields from IMERG to document the possible space–time evolution of the rainfall field during the life cycle of WEs. Four spatial types are identified. Spatial type 1 gathers the most frequent, less intense ones. Spatial type 2 is far less frequent but pictures westward-moving rainfall patterns, probably associated with mesoscale convective systems. Two spatial types (3 and 4) are related to high-intensity near-stationary rainfall events, respectively, located to the southwest and northwest of the stations. ST 4 is mostly present at stations close to the Atlantic Ocean.
Read full abstract