Weather data-centric prediction of maize non-stressed canopy temperature in semi-arid climates for irrigation management.

Canopy temperature (Tc) measurements are increasingly being used to compute crop thermal indices for water stress estimation and improved irrigation management. Conventionally monitoring crop thermal response requires maintenance of a well-watered crop from which non-stressed canopy temperature (Tcns) is measured as a reference for thermal index computation. This study alternatively evaluated the performance of 36 weather data driven model combinations to predict peak time (12:00–17:00 h) Tcns in maize grown in semi-arid climates at the West Central Research, Extension, and Education Center (WCREEC) in North Platte, NE, and at the Limited Irrigation Research Farm (LIRF) in Greeley, CO. Data-driven models considered were multilinear regression (MLR), forward feed neural network (NN), recurrent neural network (RNN), multivariate adoptive regression splines (MARS), random forest (RF), and k-nearest neighbor (KNN). For each of these models, the following weather data combinations were tested: average air temperature (Ta), average relative humidity (RH), wind speed (U2), and solar radiation (Rs) (combination 1); RH, U2, Rs (combination 2), Ta, RH, Rs (combination 3); Ta, RH (combination 4); RH, Rs (combination 5); and Ta, Rs (combination 6). Ranking the performance of weather data × model combinations across both climate sites showed that MARS model with combination 1 was a better predictor of Tcns with R2 of 0.866 and RMSE value of 0.966 °C at WCREEC and R2 of 0.910 and RMSE value of 0.693 °C at LIRF. The performance of site specific (localized) and generalized model combinations was compared and indicated that cross site prediction of Tcns was primarily determined by weather data combinations, rather than model specificity. [ABSTRACT FROM AUTHOR]