Abstract
A reliable estimate of crop water requirement is one of the key design criteria required for planning, designing and operation of irrigation schemes. This, in turn, requires accurate estimtion of reference crop evapotranspiration for which a variety of equations are widely used. The sensitivity analysis of five selected methods was conducted. The senstivity coefficients were computed by approximating the derivatives by finite differences. A sensitivity coefficient (relative sensitivity)AETo/AXi • Xi/ETo was calculated to compare the relaive importance of different input variables. The sensitivity coefficient defines the change in computed ETo due to change in the variable. The mean annual sensitivity coefficients were determined by applying 10 years data (1974-1983) obtained from agrorneteorological stations of Thailand. The sensitivity analysis indicates that the Penman method provides more realistic estimates of the importance of the various climatological factors. The Jensen-Haise method is more sensitive to temperature as compared to other methods. Based on temperature, the Jensen-Haise is the most sensitive equation and the Radiation method is the least sensitive. Between these two extremes, in the order of most to least sensitive, are the Blaney-Criddle and the Penman equations. The results also show that there is less variation in the sensitivity coefficient values of temperature in Thailand and high variation for the values of wind velocity
