Simulation of karst hydrological processes using GM(1,1) metabolic model

In this paper, the GM(1,1) metabolic model was applied to simulation of the spring discharge of Niangziguan Springs. We identified the models by error analysis and found that the GM(1,1) metabolic models with data sequence of seven and eight dimensions can reflect the future variation of the springs flow. The results showed that the predicted spring flow of the models with data sequence of seven dimensions were larger than those of eight dimensions. The models provided with a range of the spring discharge. The Niangziguan Springs flow was supposed to be declining. The upper limit of Niangziguan Springs flow was likely to decrease from 6.91 to 6.73 m³/s during 2008 to 2012. Accordingly the lower limit was likely to decrease from 6.83 to 6.57 m³/s during the same period of time. The continuous decaying of Niangziguan Springs flow stressed that karst groundwater exploitation and economic activities disturbed groundwater circulation that have exceeded natural supplement.