Rainfall Tercile Categories
Observing which areas are drier or wetter than normal.
Historical observations are used to define a climatology of wet, average and dry conditions, otherwise known as 'rainfall tercile categories'; ideally a minimum of 30 years is used for the climatology period. Current observations are then compared with the climatology and are matched to each tercile category (determining in which category a particular rainfall amount falls). This is then used to build up a picture of which areas have been drier or wetter than normal. The tercile categories can also be compared with the equivalent results from the seasonal forecast model in order to assess how well the model is performing.
Within CSRP, tercile categories are calculated independently for two different rainfall data sets. More than one data set is analysed because the observations contain errors and comparing different data sets can give some indication of where the largest errors occur. We can have the greatest confidence in the results where the data sets agree.
The two data sets used here are 10-day averages (dekads) from the Tropical Applications of Meteorology using SATellite data (TAMSAT), and daily data from the Climate Prediction Centre Famine Early Warning System Rainfall Estimates ( FEWS-NET CPC-FEWS RFE). These data sets were selected as they use different data sources and are therefore semi-independent. TAMSAT uses data from a single satellite, calibrated with historical ground-based observations, whereas CPC RFE blends data from several satellites with real-time rain-gauge estimates. Data sets that include satellite data over Africa are useful as ground-based observations data are very sparse in this region, and using these observations alone can result in a data set with very large errors.
The latest TAMSAT dekadal rainfall estimate is updated automatically and available to view on the TAMSAT website here: TAMSAT RFE Latest.