Estimation of crop coefficients and evapotranspiration by meteorological parameters in a rain-fed paddy rice [Oryza sativa] field, cassava [Manihot esculenta] and teak [Tectona grandis] plantations in Thailand

In the present study, an attempt has been made to present empirical equations including easily available meteorological parameters for estimating crop coefficients (Kc) and actual evapotranspiration (AET). The Bowen ratio method (BREB) was used to measure AET, and the Penman-Monteith (PM) equation was employed to compute reference evapotranspiration (ETo). Kc was calculated by the ratio of AET to ETo. Measurements were carried out in three popular vegetations in Thailand, a rain-fed paddy rice field, cassava and teak plantations. We investigated the correlations between Kc and individuals as well as combinations of the meteorological parameters of solar radiation (Rs), air temperature (Ta), wind speed (WS), vapor pressure deficit (VPD), and soil water content (SWC) in each site. Moreover, the above-mentioned correlations were examined using pooled data of the paddy and cassava as well as combined data of all sites. The highest correlation coefficients were found between Kc and combinations of the five mentioned parameters for the three sites. We found that the correlation coefficients between Kc and four easily available parameters excluding SWC were nearly the same as those of the five parameters. Since SWC data are sometimes unavailable, the authors presented the equations including only the easily available parameters to obtain the optimum estimation of Kc. One equation was obtained from the pooled data of paddy and cassava. Root mean square error (RMSE) of the estimating AET using this equation was 0.84 mmday-l on a 1-day scale. The authors finally proposed another equation for the three sites. RMSE of the final equation on a 1-day scale was found to be 0.72 mmday-l that corresponds to 19% error. Accordingly, errors of the estimated AET by the final equation at 5, 10, 15 and 20-days of averaging time scales were 15, 12, 11 and 10%, respectively. Therefore, we concluded that averaging over more than 10 days might be more suitable for estimating AET by the final equation.