The crop coefficients of winter wheat and summer maize in the plain and mountain areas in the middle area of Shandong Province were modified by soil moisture balance equation, then the theoretical model and statistical model were established, and the backward substitution and prediction analysis were conducted.The results showed that the crop coefficients of winter wheat and summer maize in plain were higher than those in mountain area in the middle area of Shandong Province, and their peak values appeared at jointing-milky stage and jointing-tasseling stage respectively.The absolute root mean square errors of backward substitution and prediction were within 5 mm in most time of the two models, and their relative root mean square errors were basically less than 2%.The backward substitution and prediction results of two models on soil moisture in the plain were close to each other, in which, the error of statistical model was slightly smaller.The backward substitution and prediction results of statistical model in the mountain area was better in most time.The prediction results of two models in the plain were better than those in the mountain area.Thus, the statistical model could be used in the actual business, but its stability and mechanism still needed further research.%通过土壤水分平衡方程对鲁中地区平原和山区两种地形冬小麦和夏玉米的作物系数进行修正,建立了理论模型和统计模型,并进行了回代和预报对比分析.结果表明:鲁中地区冬小麦、夏玉米生育阶段作物系数山区较平原低,最大值分别出现在拔节至乳熟、拔节至抽雄阶段;两种模型多数时段回代和预报绝对均方根误差均在5 mm内,相对均方根误差基本在2%以内;两种模型对平原地区土壤水分的回代和预报效果接近,统计模型误差略偏小,对山区的回代和预报效果多数时段统计模型偏好;模型对平原土壤水分的预报效果较山区好.综上所述,统计模型基本能用于实际业务中,但其稳定性和机理机制仍需深入研究.
展开▼
