Considering the classification assessment of water quality can not fairly reflect the over pollution degree of the water
body, this article puts forward a measurement on over pollution based on kernel density estimation, to evaluate water
quality of the main stream of Weihe River with the monitoring data of its rainy, dry and normal seasons from 2000 to
2008. Through the comparison between the assessment result by a Standard Category Index Method, and the standard
of water function zoning plan, the space distribution of water function zoning is found to be inconsistent with that of
status quo of water quality, therefore, from the spatial dimension, the method of PCA-CA is used to classify the water
quality situation of 13 existing monitoring sites, and according to ANOVA, the cluster result can significantly
distinguish the difference between the actual land use of waterfront; from the time dimension, rainy season is
comparatively different from dry season and normal season with these three seasons as time standard which is often
used to differentiate the quantity of flow. This article discusses the characteristic of water quality overload of Weihe
River in Shaanxi with the change of seasons, and different clusters of area, also, concludes the over pollution extent
of three areas in different seasons based on the classification result from the space and time dimension. The result
shows that the biggest overload extent happens in Area1 and Area3 during dry seasons, and Area 2 during normal
seasons; overally considering with the 4 indicators, NH3-N and BOD5 are the most seriously polluted indicators in
Weihe river.