Optimal management of cultivated land coupling remote sensing-based expected irrigation water forecasting

This study proposed a novel cultivated land planning approach coupling expected irrigation water prediction and crop planting structure optimization. Under the guidance of the expected irrigation water prediction, the development of agricultural planting strategies was proposed through the cultivated land planning method to improve the efficiency of water use and ensure agricultural production. This approach was consisted of two models: a combination forecasting model was developed for estimating expected irrigation water with help of remote sensing data (RSCM); and a bi-level multiobjective programming model for planning crop planting structure (BMPSOM) was formulated for crop planting structure programming. Then, the proposed approach applied to a real-world case in Northwest China for verifying its validity. Results show that: 1) the forecast accuracy of the combination forecast model has been greatly improved compared to the single forecast model, the accuracy test results are RMSE = 0.25 & times; 108 m3, MAPE = 0.78%; 2) the use of remote sensing data can reflect the soil moisture with spatial variability in the forecast; 3) BMPSOM model can well reflect the leader-follower relationship and make tradeoff among multiple objectives of each decision maker in tackling practical problem. Application has proved that the agricultural planting strategy formulated by this approach can not only ensure economic benefits but also increase food production under limited water resources. Therefore, this method has good application prospects in areas where water resources are scarce.