Regionalization and spatial estimation of ethiopian mean annual rainfall

摘要: This study aims at developing methods that will lead to the generation of a reliable mean annual precipitation data base for Ethiopia. Multiple regression models have been formulated explain rainfall as function elevation and geographical location. The estimations, based on yearly values from set 63 Ethiopian stations with records between 1969 1985, were developed whole country well already existing Food Agricultural Organization (FAO) pattern regions new zonation derived by principal component common factor analyses (PCA/CFA). In PCA/CFA study, monthly 1968 1985 43 used. optimal was testing 36 different combinations resulting in regions. alternatives tested were: correlation covariance dispersion matrices, PCA CFA eigentechni-ques, unrotated rotated (Varimax Direct Oblimin) components/factors number possible significant (3, 7, 11). Principal analysis matrix, Varimax rotation seven extracted components gave far best relationship rainfall, elevation, Models explaining least 72 per cent variation constructed covering about 98 country, which is better than FAO (69 explained 81 country) model (66·5 variation).