MODELING CLIMATE FOR ECOLOGICAL STUDIES IN SOUTHERN KENYA

Abstract

Ecological studies often use Geographic Information System (GIS) models to predict or explain biogeographicalpatterns and range distributions of various species. Climatic data are often interpolated into high resolution spatialGIS layers to be used in GIS models. The wildlife rich Southern Kenya, generally characterized by two dry seasonsand two wet seasons, has a sparse and irregularly spaced network of weather stations, presenting a challenge forecologists who intend to use climatic data in their models. The suitability of Regression Analysis (RA) to interpolateclimatic surface models for this region at various temporal scales was examined using analysis of variance(ANOVA). It was found that for all but one wet season, precipitation could be accurately modeled using RA.ANOVA results also indicated that it was possible to accurately model temperature at various temporal scales.