Piecewise analysis and prediction of spring flow based on GM(1,1) model

One of the largest springs in North China, the Liulin Karst Springs, has been decaying since 1950s. We introduced the piecewise analytical concept to GM(1,1) model focusing on analysis and prediction of Liulin Karst Springs discharge. We divided the spring discharge processes into two phases: pre-1973 and post-1973. In the first phase (i.e. 1957-1974) the spring discharge was affected by climate change and in the second phase (i.e.1973-2007) the spring discharge was influenced by both climate change and anthropogenic activities. In the first phase, spring flow was simulated by the grey system GM(1,1) decomposition model. Then the effects of climate change were modeled. By extrapolating the model, we acquired a calculation of the spring discharge in the second phase sole under the effects of climate change. Based on water balance analysis, we conclude that the contribution of climate change to depletion of Liulin Springs is 0.18-0.34 m³/s from 1974-2007. However, the contribution of anthropogenic activities to depletion of spring flow is 2.55-2.70 m³/s in 2007. The contribution of climate change only accounts for 10% of contribution of human activities. Therefore, human activities have large contribution to Linlin Spring depletion. Regarding to the contribution of human activity effects on spring discharge decline, groundwater pumping only account for 20%-35% of the spring flow depletion; 65%-80% of the depletion are likely to be caused by other human activities including dam-building, dewatering during coal mining, and deforestation.