Research On Soil Moisture Prediction Models Of Inadequate Irrigation Rice Field In South China

The soil moisture refers to the soil water status in the main root system activity layer of crops. Soil moisture prediction is one of the key techniques of the dynamic distribution of irrigation system in the irrigation district. The rice is one of the main food crops of our country. The stability and development of the production scale of the rice are very important in grain-production. The water that the rice needs is more than others. With the development of the saving irrigation and the pinch of water resource more and more, it is very important to research the upland rice cultivation and inadequate irrigation. Predicting soil moisture of the rice field and high-efficient management of irrigation have great theoretical and practical meaning to keep continuously water and soil resources and develop of sustainable agricultural production.In order to handle the shortcoming of the soil moisture of inadequate irrigation rice field prediction the paper sets up an empirical model and an artificial nerve network model to forecast soil moisture by field trail, interior simulation cultivation and computation with the mathematical model.Then the paper analyses the growth of the rice food system in tillering stage. Based on the substantive data, we establish the rice root water uptake model and the macroscopical model of water dynamics of soil.The packets in the soil water forecast are compared with the packets in observation. The research indicated the neural network model is the most accurate. The experience model takes the second place and the water dynamics model is relatively lowest. These may relate to experimental means and precision of equipments.