Abstract: The dynamics of soil water content during plant growth peried can be decomposed into two parts:periodic term and random fluctuation term. In this study, simulation treatment of time sequence wasearried out by using method of harmonic analysis to periodic term whereas linear regression was established to random fluctuation term vs. Natural precipitation by using obervation data of soil water contentduring 1992. Predictive simulation equation of dynamics of soil water content during 1992. Predictive simulation equation of dynamics of soil watent content in alpine meadow was set up by using plus modal method. Regressive simulation was used for the date of soil water content of 1992. The predictive imitation the was very high. The mean relative error is 7.2%. prediction for the dynamics of soil water content of1993, the accurate rate was very high the, and it’s mean relative error is 8.9%. This result showed that the simulant model equation can predict the dynamics of soil water content accurately, and it also can guide the livestork farning. So the study on the ecology of eneng cycle of grassland matter has a great actual price.

Key words: Alpine meadow, Dynamics of soil water content, Simulation, Harmonic analysis

摘要： 土壤水分在植物生长期间的动态变化,可分解为周期项和随机波动项两个部分。利用1992年土壤水分观测资料,以谐波分析法对周期项进行随时间序列的模拟处理,以剩余土壤水分(随机波动项)建立与自然降水量的线性回归关系后,依加法模式确定高寒草甸地区土壤水分动态变化的预报模型方程为: 对1992年土壤水分资料进行回代模拟,拟合率很高,平均相对误差为7.2％。对1993年土壤水分动态变化进行预测预报,准确率较高,其相对误差平均为8.9％。结果表明,该模拟方程,能准确预报土壤水分动态变化的过程,进而对研究草地物质能量循环等生态过程均具有一定的实用价值。

关键词: 高寒草甸, 土壤水分动态变化, 模拟, 谐波分析