Can complementary methods reliably estimate evapotranspiration in semi-arid regions?

Document Type


Journal/Book Title/Conference

Spring Runoff Conference

Publication Date



In semi-arid regions, the largest portion of rainfall is lost as evapotranspiration (ET). Therefore, the uncertainty in estimating ET can lead to the inaccurate prediction of water balance and water resources needs. Furthermore, rural river basins are characterized by scarcity of data and resources, adding additional challenges in estimating water resources needs in the basin. This research aims at estimating ET in semi-arid rural river basins with limited data for the purpose of water resources planning and management. In the literature, several classical methods are used to estimate potential ET. However, estimating actual ET requires detailed local data such as temperature, land cover/land use, crop pattern, growing cycle, etc. Still, these classical methods are mainly applicable to predict ET from crop covered areas during the growing seasons. On the other hand, actual water loss from the land surface is not typically restricted to crop areas only; instead evaporation could happen from open water bodies as well as from open land surfaces with minimal vegetation cover. In water resources planning, the important estimate needed is the total water loss from the land surface that may or may not include transpiration from crop areas. For several decades, complementary methods, including Complementary Relationship Areal Evapotranspiration (CRAE), Advection-Aridity (AA), and Granger and Gray (GG) methods, have been used to estimate ET. These methods are attractive due to simplicity, practicability, and reliability in estimating actual, wet environment, and potential ET at regional scale using meteorological data only. Previous studies attempted to use the complementary methods with little success given the limited understanding of the methods and the confusion due to the definitions of various terms. Still the complementary methods offer a distinct advantage over the classical method given the simplicity of data and the ability to estimate total water loss as opposed to ET. The purpose of this study is to investigate the applicability of the complementary methods in estimating ET and the needs to perform additional revisions to the methods to improve estimates if necessary. A rural semi-arid region in northern Ghana is used as a case study. The monthly meteorological data from five synoptic stations are used in the analysis. The results show that the complementary methods overestimate ET. The probable reason for this discrepancy was due to the over estimation of net radiation. Once the method was modified, the estimates were promising compared to the results from prior studies conducted in the same region. It is found that the complementary wet environment ET is close to the Penman estimates while the complementary potential ET estimate is close to pan evaporation. The long term annual mean rainfall ranges from 984 to 1241 mm while the long term annual mean ET is 64% to 80% of rainfall.

This document is currently not available here.