中文摘要：

在气候干旱的宁蒙引黄灌区，控制地下水位是防治土壤盐碱化、保证农业健康发展的重要途径。采用相关分析的方法确定了影响灌区地下水位变化的主要因素，建立了基于LM算法的灌区神经网络地下水位预测模型，并以宁夏河东灌区为实例进行了研究。研究结果表明：模型能够较好地模拟灌区地下水系统的变化特征，对地下水位做出较准确的预测。

英文摘要：

Controlling the groundwater table is very important to preventing soil salinization and agricultural development in Yellow fiver irrigation areas. In this article , major factors influencing the groundwater table of irrigation are analyzed by the correlation method, and a neural network groundwater table prediction model is built based on LM algorithm. The Ningxia Hedong irrigation area is conducted as an example research, it shows the groundwater system can be well simulated by the model, and the groundwater table can be accurately predicted by the model.