Temperature-Driven Models for Estimating Reference Evapotranspiration in Semi-Arid Malawi

Abstract

The importance of evapotranspiration in agricultural water management has been widely reported without doubt. The 
FAO Penman Monteith (FPM) equation is the sole recommended estimation method of reference evapotranspiration 
(ETo), but its application is limited due to large number of meteorological data required. In such circumstance, 
Hargreaves (HGR) and Blaney-Criddle (BCR) temperature-based empirical models are still used in Malawi, regardless of 
their accuracy in some semiarid areas. This study explored the potential of using HGR and BCR temperature-based 
models, in Ngabu and Chileka, located in Southern Malawi, where decade climatic data required for FPM equation was 
collected from 1985 to 2004. The data sets were divided into dry and rainy season to take into account the tendency of 
over and under estimation of ETo by the models. It was found from this study, that splitting the data into dry and rainy 
season improves the accuracy of the temperature-based models. From the statistical comparison with FPM, in both sites, 
HGR performed better than BCR during dry season while BCR showed superior performance over HGR during rainy 
season. Therefore, seasonal consideration is strongly recommended when temperature-based models are applied under 
semiarid climatic condition like Malawi where radiation (r = 0.87) and wind speed (r = 0.89) affects ETo. Finally, seasonal 
temperature- based models are determined and recommended in this study to be used as alternatives of FPM for ETo 
estimation in Ngabu and Chileka of Malawi.