Analysis On Dynamic State And Prediction Of Groundwater In Arid Oasis

Oasis is the core area of arid region and the major area of human production,industry and agriculture.Yaoba oasis,located in the west of the Helan Mountain and the east of the Tengger Desert,is one of the top ten desert oases in Alxa League of Inner Mongolia,which is the largest agricultural base in Alxa League as wel.Due to the lack of surface flow and rainfa ll,groundwater resources become the main source of water supply in this region.With the continuous development in recent years,the increasing population and unlimited exploitat io n led to rapid declining of the groundwater level in Yaoba oasis,resulting in a series of ecologica l crisis,which seriously affected the local people's living,economic development and ecologica l security.Therefore,exploring the dynamic change and trend of the groundwater level under the influence of climate change and human activities has a significant guiding impact for rational utilization of groundwater,which is directly related to the sustainable development of social economy in Yaoba Oasis.In this paper,the groundwater level of Yaoba Oasis was taken as the research object.Firstly,the dynamic characteristics and cycle evolution law of the groundwater level was analyzed.The second objective is to quantitatively discuss the influence of climate changes and human activities on the groundwater level.By constructing groundwater level forecasting model,the final objective is to simulate the trend of the groundwater level in the future.The main research results of this paper are as follows,1.Basing on the analysis of the groundwater level's annual and interannual variatio n characteristics and the development of the cone of depression,this paper uses the Morlet wavelet transform to analyze the periodic evolution rule.The results showed that the annual variation of the groundwater level had distinct seasonal characteristics,which was divided into5 stages of the spring irrigation declining period,spring irrigation intermittent recovery period,summer irrigation declining period,autumn irrigation recovery period and winter irriga t io n recovery period from 1981 to 2016.The interannual variation of the groundwater level was divided into 4 stages of the slowly declining period,sharply declining period,slowly declining period and stable period.The cone of depression of the groundwater level undergone a process of formation,expansion and reduction.The time series of the groundwater level which the first principal period was 12 months,mainly existed two time scales of 17-35 months and 7-15months.2.The linear trend analysis and the Mann-Kendall test are used to estimate the trend and test the abrupt change of the climatic factors from 1954 to 2016.This paper found that the temperature which was consistent with the global warming,had a significant upward trend,the rainfall increased insignificantly,the evaporation and relative humidity decreased significant ly.And nearly 60 years,the usage of the groundwater resource increased due to the human activities(irrigated area expansion,population growth and exploitation quantity increasing),which resulted in the reduction of the groundwater level.Therefore,climatic factors and human activities resulted in a significant decrease of the groundwater resource in Yaoba Oasis.3.The nonlinear relationship between the groundwater level and driving factors is quantitatively analyzed.This paper used the principal component regression to find that the human factors accounted for 86.65%and climate factors accounted for 13.35%,so the human activities played a leading role on the groundwater level variation,especially the human exploiting.However,the rainfall and the evaporation had a small influence on the groundwater level variation4.Respectively using the ABC-BP neural network model and BP neural network model to train and simulate the data of the groundwater level,this paper finds that the accuracy,speed and stabilization of the ABC-BP neural network model is better than the BP neural network model.Using ABC-BP neural network model which has the double hidden layers to predict the groundwater level of 2018-2030 under different mining modes,the results show that the groundwater level is in a declining stage under the existing mode,the groundwater level is in a stable stage when the exploitation quantity is adjusted to 31million m~3/a,and the groundwater level is in a recovery stage when the exploitation quantity is adjusted to 22million m~3/a.