Three methods of estimating the power of maximum temperature in TM–ET estimation equation

Abstract

The use of FAO56-PM is still the preferred method for the estimation of evapotranspiration (ET) regardless of locations. But in places where data that satisfy all the meteorological variables of PM equation are non-existent or missing, the use of either temperature or radiation-based methods is the alternative option, since both radiation and temperature are the main drivers of ET. From the latter two, temperature-based ET estimation methods are still the simplest ones to use. However, there are challenges since some of the parameters used in the temperature-based equations are location dependent. Such location dependence is due to variability in terms of latitude, altitude and other geo-location parameters. Therefore, local calibration of temperature-based methods is necessary to get better estimates of ET. In this study, one temperature-based ET estimation method known as Temesgen-Melesse's (TM) method was assessed in relation to the PM equation using data of ten Class I meteorological stations in Ethiopia. In the study, three different techniques were tested to find location-dependent maximum temperature power ('n') used in TM equation (i.e., to calibrate the method). The first two techniques involve calibration using PM data. In the first case, monthly averaged data that span from 1 to 5 years were used (as five data sets of 12, 24, 36, 48 and 60 months). In the second case, daily values that span from 10 to 30 days were used in three groups of 10-, 20-, and 30-day data sets. In the third technique, an attempt was made to estimate 'n' from the location latitude, altitude and average of the monthly averaged maximum temperature (T̄m x ). The results obtained from each technique are given as follows. Calibration from the monthly averaged data gave good 'n' values for all the stations with R 2 values ranging from 0.80 to 0.92. There were no differences in the number of data points (12,24, 36, 48 and 60 months) used, which means data points of 12 months are sufficient for the calibration. Calibration using the daily data gave satisfactory 'n' results for all the stations tested, though the statistical parameters and performance tests did not give results that are comparable to the monthly averaged values. The results were nearly the same for 10-, 20-, or 20-day data points, which means any one of them can be used for calibration purposes. Estimation of 'n' from latitude, altitude and T̄m x gave results that are comparable to the ones with 'n' calibrated. The percent differences between the 'n' calibrated using data of 131 months and the 'n' values obtained from monthly data and daily data calibrations are less than 0.16% and 0.5%, respectively. The calculated 'n' showed a percent difference of less than 1.1%. Using calculated 'n' is better than performing calibration using daily data. It can also be considered as option when at least 1-year PM data are not available for calibration.