Projection Pursuit Method for Analysis of the Relationship between the Precipitation and the Runoff from Heilongtan Spring in Lijiang Basin of Yunnan Province, China

Projection pursuit regression is a statistical model which adapts the additive models in that it first projects the data matrix of explanatory variables in the optimal direction before applying smoothing functions to theses explanatory variables. This paper mainly discusses the projection pursuit regression (PPR) model developed for analysis of the relationship between the precipitation and the runoff from Heilongtan spring in Lijiang basin of Yunnan Province of Southwest China. A long series precipitation data of four rainfall stations and the spring flow monthly data from year 2001 to 2011 are used for establishing of PPR model. The results illustrate that the spring flow is lagging behind the rainfall about 120days, and is much significantly relative to the precipitation of Lijiang rainfall station which is located in plain area and near the springs, instead of the precipitation of Jiuzihai, located in recharge area, higher mountain area. This is because the variation of precipitation in plain area causes the change of groundwater level of pore aquifers in plain areas, and affects the spring flow of karst groundwater which directly hydraulic-connected with pore aquifers in plain area. The precipitation in recharge area also affects the variation of spring flow of karst groundwater, but not very significantly, because of the regulation of huge storage of karst aquifers. The research results could be very meaningful for the scientific development and planning of water resources as well as the protection of the karst springs in Lijiang basin.