农田耕作土壤田间持水率的BP预测模型

BP Prediction Model Tillage of Soil Water Content at Field

为了合理确定疏松黄土的灌溉制度,基于黄土高原区农田耕作层土壤田间持水率的测试资料,建立了以土壤理化参数包括土壤体积质量、黏粒量、粉粒量、有机质量为输入变量,田间持水率为输出变量的BP预测模型。结果表明,预测值和实测值的相对误差控制在7%以内,在可接受的范围,表明利用土壤基本理化参数预报农田耕作土壤的田间持水率是可行的。研究结果可为黄土高原区耕作农田土壤水分调控和灌溉管理提供支撑。

In order to determine the loose loess irrigation system reasonably, based on the Loess Plateau Soil Till- age field test data of holding water rate, a BP model was established, using soil physical and chemical parameters including soil bulk density, clay content, silt content, organic matter as the input variables, to forecast the field holding water rate. The results showed that the relative error between predictive value and the measured value was controlled within 7%, which was in the acceptable range, indicating that using soil basic physical and chemi- cal parameters to forecast the tillage field soil water retention rate was feasible. The results could provide support for controlling the Loess Plateau Soil Moisture cultivated farmland irrigation management.