Eight radiation-based equations for determining evaporation were evaluated
and expressed in five generalized forms. Five evaporation equations (Abtew,
Hargreaves, Makkink, Priestley and Taylor and Turc), where each represents
one generalized form, were then compared with pan evaporation measured at
Changins station in Switzerland. The comparison was first made using the or
iginal constant values involved in each equation, and then using the recali
brated constant values. Evaluation of the Priestley and Taylor equation req
uires net radiation data as input, in this study, net radiation was estimat
ed using Equation (16) owing to the lack of observation data. The results s
howed that when the original constant values were used, large errors result
ed for most of the equations. When recalibrated constant values were substi
tuted for the original constant values, four of the five equations improved
greatly, and all five equations performed well for determining mean annual
evaporation. For seasonal and monthly evaporation, the Hargreaves and Turc
equations showed a significant bias, especially for cold months. With prop
erly determined constant values, the Makkink and modified Priestley and Tay
lor equations resulted in monthly evaporation values that agreed most close
ly with pan evaporation in the study region. The simple Abtew equation can
also be used when other meteorological data except radiation are not availa
ble. Copyright (C) 2000 John Wiley & Sons, Ltd.