Modeling Rice Evapotranspiration with Partial Least-Squares Regression

Calculating the evapotranspiration with weather data, it was found that some independent variables had interactions with each other. This phenomenon can distort and destabilize the multivariate regression model of traditional least square method. The partial least-square regression for model was applied to model the rice evapatranspiration base on the main component analysis and typical correlation analysis of data. The model of rice evapotranspiration was suggested to solve the teractive correlation among the independent variables (weather parameters.). The model was found to be able to give satisfactory predictions.