Wet and dry spells in Senegal: comparison of detection based on satellite products, reanalysis, and in situ estimates

Abstract

Listen

Translate article

Abstract. In this study, the detection and characteristics of dry/wet spells (defined as episodes when precipitation is abnormally low or high compared to usual climatology) drawn from several datasets are compared for Senegal. Here, four datasets are based on satellite data (TRMM-3B42 V7, CMORPH V1.0, TAMSAT V3, and CHIRPS V2. 0), two on reanalysis products (NCEP-CFSR and ERA5), and three on rain gauge observations (CPC Unified V1.0/RT and a 65-rain-gauge network regridded by using two kriging methods, namely ordinary kriging, OK, and block kriging, BK). All datasets were converted to the same spatio-temporal resolution: daily cumulative rainfall on a regular 0.25∘ grid. The BK dataset was used as a reference. Despite strong agreement between the datasets on the spatial variability in cumulative seasonal rainfall (correlations ranging from 0.94 to 0.99), there were significant disparities in dry/wet spells. The occurrence of dry spells is less in products using infrared measurement techniques than in products coupling infrared and microwave, pointing to more frequent dry spell events. All datasets show that dry spells appear to be more frequent at the start and end of rainy seasons. Thus, dry spell occurrences have a major influence on the duration of the rainy season, in particular through the “false onset” or “early cessation” of seasons. The amplitude of wet spells shows the greatest variation between datasets. Indeed, these major wet spells appear more intense in the OK and Tropical Rainfall Measuring Mission (TRMM) datasets than in the others. Lastly, the products indicate a similar wet spell frequency occurring at the height of the West African monsoon. Our findings provide guidance in choosing the most suitable datasets for implementing early warning systems (EWSs) using a multi-risk approach and integrating effective dry/wet spell indicators for monitoring and detecting extreme events.