Sensitivity Analysis and Uncertainty Parameters Affecting in The Estimation of Reference Evapotranspiration in Models With Different Mathematical Structure

Evapotranspiration is of fundamental components of water balance computations to estimate crop water requirement. To calculate this component there are a lot of direct methods which are expensive and time consuming. Many experimental methods such as FAO Penman-Monteith have been presented to calculate reference crop evapotranspiration. All ETo estimation models with different mathematical structure relative to meteorological parameters have different sensitivities and standard errors in different climates. Hence، it is necessary to analysis uncertainty in mathematical structures of these models. In this study، in addition to sensitivity analysis of seven models of estimating ETo relative to climatic parameters in Kerman station، also standard deviation value of ETo was obtained based upon variance and covariance of parameters and first-and second-order partial derivatives of ETo than these parameters. Results showed that data uncertainty in the cold seasons was more than warm seasons and all elected models had more sensitive to two parameter of Tmax and Tmin. Moreover، the maximum correlation coefficient was obtained +0. 52 which belongs to pair parameters of RHmax and RHmin and minimum one was +0. 003 which belongs to pair parameters of U2 and Tmax. Range of standard deviation of ETo for FAO Penman-Monteith، Modified Blaney- Criddle، Prestly-Taylor، Hargreaves-Samani، Hargreaves، Irmak-Rs، Irmak -Rn were achieved 0. 78-5. 49، 2. 43-8. 5، 1. 75-8. 61، 0. 18-1. 03، 0. 19-1. 09، 0. 16-0. 96، 0. 15-0. 56 mm d-1 during a year. Based on data in this research، ∆ETo is enormous (-480 mm d-1) in cold weather seasons when minimum temperature tends to zero