Grey system analysis software development for karst ground water modeling

After analyzed recent application of grey system theory in karst hydrology, a grey system modeling software is developed by means of step wise refinement method and VisualC++ 6.0 with the thoughts of software engineering. There are six modules in the software including grey relational analysis, data pretreatment, grey relational analysis, GM(1,1) model, dimension-fixed and recursion-compensated prediction model, GM(1,1) decomposition model, and GM(1,N) model as well. It is object-oriented and easy to operate, and able to meet the needs of hydrologists for groundwater simulation. This paper describes the software's design process, analyzes the functions and implementation of each module, presents the software examination process, and states the software's benefits on the basis of a case study in Liulin Springs, Shanxi, China. AGM(1,1) model is set up for Liulin Springs in light of the spring discharge data from 1971 to 2007, and the model is verified by spring discharge data in 2008 and 2009. The results show that the residuals errors decrease from 0.32 m3/s to 0.17m3/s in 2008 and from 0.29m3/s to 0.19m3/s in 2009 by periodic correction to the residuals.