Application of sequential clustering method in vertical dividing of soil moisture profiles

In modern researches on soil moisture dynamics, the methods of dividing soil moisture profiles have no rigorous mathematical theory to support.To resolve this omission, we carried out field monitoring in 2005 to the depth of 200 cm on four Robinia pseudoacacia sites in Ji County, western Shanxi Province of the Loess Plateau.Considering that clustering results of sequential clusters did not change the sequence of samples, we used this method to divide the soil water profile of four R.pseudoacacia sites with the data of soil moisture obtained from a fixed field monitoring system.The results indicate that the soil moisture profile of R.pseudoacacia in this area can be divided into three layers:weak moisture utilizing layer(0-20 cm), moisture utilizing layer(20-100 cm) and moisture regulating layer(below 100 cm).This is in almost complete agreement with the results of other researches.The results show that the seguential clustering method has a rigorous mathematical theory foundation.Combining with the sequence of the soil moisture profile, this method is an innovation in the research of dividing soil moisture profiles.