PREDICTING THE DISTRIBUTION OF TSETSE-FLIES IN WEST-AFRICA USING TEMPORAL FOURIER PROCESSED METEOROLOGICAL SATELLITE DATA

An example is given of the application of remotely-sensed, satellite d ata to the problems of predicting the distribution and abundance of ts etse flies in West Africa. The distributions of eight species of tsets e, Glossina morsitans, C. longipalpis, G. palpalis, G. tachinoides, G. pallicera, G. fusca, C. nigrofusca and G. medicorum in Cote d'Ivoire and Burkina Faso, were analysed using discriminant analysis applied to temporal Fourier-processed surrogates for vegetation, temperature and rainfall derived from meteorological satellites. The vegetation and t emperature surrogates were the normalized difference vegetation index and channel-4-brightness temperature, respectively, from the advanced, very-high-resolution radiometers on board the National Oceanic and At mospheric Administration's polar-orbiting, meteorological satellites. For rainfall the surrogate was the Cold-Cloud-Duration (CCD) index der ived from the geostationary, Meteosat satellite series. The presence o r absence of tsetse was predicted with accuracies ranging from 67%100% (mean=82.3%). A further data-set, for the abundance of five tsetse sp ecies across the northern part of Gate d'Ivoire (an area of about 140 000 km(2)), was analysed in the same way, and fly-abundance categories predicted with accuracies of 30%-100% (mean=73.0%). The thermal data appeared to be the most useful of the predictor variables, followed by vegetation and rainfall indices. Refinements of the analytical techni que and the problems oi extending the predictions through space and ti me are discussed.