Dynamic Change Of Groundwater Level In Minqin Basin Based On Various Methods

Groundwater is the natural resource for human survival and development.Groundwater level dynamics is the external expression of groundwater balance.Scientific demonstration of groundwater level dynamic change law and study of groundwater dynamic analysis and simulation methods can provide useful tools for scientific management of groundwater and have important significance for sustainable development of groundwater resources.Groundwater is an important source of water supply for production and life in Minqin Basin of China,and it also maintains a fragile ecological environment,which is an extremely important environmental factor.Based on time series,multivariate regression and neural network methods,taking Minqin Basin as the research area,this paper carries out the prediction of groundwater dynamic change on the basis of in-depth analysis of hydrogeological conditions in Minqin area,which provides scientific basis for rational optimization of water resources allocation.The main conclusions are as follows:(1)Three time series forecasting methods,i.e.factor combination decomposition forecasting method,Holt-Winters exponential smoothing method and ARIMA model,are used to construct the groundwater level dynamic change model of monitoring holes in Huanhe Water Pipeline Station.By comparing the actual value and forecasting value,it is concluded that the mean square error(MS)of ARIMA model is the smallest and the fitting degree of the model is good,which can be used in the groundwater dynamic forecasting Based on this method,the dynamic changes of groundwater level in Huanhe Water Pipeline Station in the next five years are predicted.(2)By collecting the relevant data of discharge and recharge items in Minqin area,the multivariate linear regression analysis model and the artificial neural network model are established.By comparing the predicted value with the actual value,it is concluded that the mean square deviation(MS)of the artificial neural network model is smaller than that of the multivariate linear regression model and the fitting degree is good,which can be used in the study of groundwater dynamic prediction in this area.(3)The trend of groundwater table dynamic change in Minqin Basin mining groundwater dynamic type area shows a fluctuating decline,which reflects the characteristics of mining groundwater dynamic type in the basin.With the passage of time,the downward trend of groundwater level remains unchanged,but the downward trend slows down.The forecast trend of groundwater level in the next five years will be flat as a whole,indicating that the groundwater level of exploitation type will be in a stable state.The groundwater level of the infiltration-exploitation groundwater dynamic type area fluctuates before 2010.After 2010,due to the implementation of relevant regulations and the increase of upstream water inflow,the groundwater level shows an upward trend as a whole.The groundwater level in the five-year forecast keeps rising,which indicates that the groundwater level in this area will be in a continuous upward state.