Evaluating the Performance of Calibrated Temperature-based Equations as Compared to Standard FAO-56 Penman Monteith Equation in Humid Climatic Condition of Dehradun (India)

Abstract

The present study was undertaken to calibrate, validate, and evaluate the performance of temperature-based evapotranspiration equations in comparison to the standard FAO-56 Penman Monteith (FAO-56 PM) model in humid climatic condition of Dehradun district of Uttarakhand using 31-years (1989-2019) daily meteorological dataset. The quality control of dataset was ensured by omitting days with missing data and outliers. The performance of 12 calibrated temperature-based ETo equations namely, Allen (1993) [AL93], Baier and Robertson (1965) [BR65], Bogawski and Bednorz (2014) [BB14], Droogers and Allen (2002) [DA02], Dorji et al. (2016) [DO16], Hargreaves (1994) [HA94], Heydari and Heydari (2014) [HH14], Kharrufa (1985) [KA85], Ravazzani et al. (2012) [RA12], Samani (2004) [SA04], Schendel (1967) [SC67], and Trajkovic (2007) [TR07] were evaluated in comparison to standard FAO-56 PM model in terms of daily ETo estimates. The analysis showed that calibrated temperature-based equations performed well with higher value of agreement index (0.85-0.98) and reduced errors. The values of Root mean square error (RMSE), Mean bias error (MBE), Maximum absolute error (MAXE), Percent error of estimate (PE), and Standard error of estimate (SEE) for calibrated equations ranged from 0.29 to 1.15 mm.day^-1, (-)0.39 to 0.53 mm.day^-1, 0.64 to 3.95 mm.day^-1, 4.71 to 19.11%, and 0.17 to 1.00 mm.day^-1, respectively; whereas for original equations they varied in the range from 0.47 to 3.64 mm.day^-1, (-)0.32 to 2.92 mm.day^-1, 0.95 to 10.65 mm.day^-1, 11.74 to 106.15%, and 0.18 to 1.72 mm.day^-1, respectively, indicating improved performance of calibrated equations. The ranking of calibrated ETo equations on the basis of Global Performance Indicator (GPI) values confirmed that calibrated Dorji et al. (2016) equation produced best result, while Samani (2004) equation with its lowest value performed poorly. Based on GPI values, calibrated equations can be ranked (best to worst performing) as DO16> AL93> TR07> RA12> DA02> HA94> SC67> KH85> BR65> HH14>BB14> SA04. Thus, calibrated Dorji et al. (2016) equation can be used as substitute for FAO-56 PM model in the absence of large number of meteorological parameters for accurate estimation of ETo values in the study area.