摘要
基于东北地区近30 a(1981—2010年)土壤湿度观测资料,采用CAST聚类方法进行土壤湿度区划,对各区域土壤湿度的时空变化特征展开分析,建立土壤湿度的分区预报模型。结果表明:东北地区中部的土壤湿度在近30 a内呈上升趋势,其余部分则呈下降趋势,但北部的下降趋势较小,西部下降趋势较大;各区域土壤湿度的显著影响因子有所差异,但都与前一旬的土壤湿度和降水量有较大的相关性;利用1981—2007年的数据建立各区域的土壤湿度预报方程,利用2008—2010年的数据对预报方程进行检验,土壤湿度的预报平均相对误差小于13.67%,预测值与观测值较为接近,基本可反映土壤湿度的实际变化情况。
The spatial-temporal variations of soil moisture in Northeast China are analyzed by using the observation data from 1981 to 2010 with cluster analysis statistic test(CAST). The regional soil moisture forecasting model is also established. The results show that the soil moisture shows an upward trend in central part and a downward trend in other parts in recent 30years, but the western part shows a larger decline, the southern part presents a smaller decline.The significant impact factors of soil moisture are different in different parts of the region. But the correlation is evident when compared with precipitation and soil moisture in ten days before. The forecast equations of different parts are established with the data from 1981 to 2007and checked with the data from 2008 to 2010. The average relative errors are all less than13.67%. The predicted value is closed to the observation value, so the real situations of soil moisture can be showed by this predicted value.
出处
《自然资源学报》
CSSCI
CSCD
北大核心
2014年第6期1065-1075,共11页
Journal of Natural Resources
基金
公益性行业(气象)科研专项(GYHY201106026)
关键词
土壤湿度
预报模型
聚类统计
东北地区
soil moisture
forecasting model
cluster analysis statistic test
Northeast China
