Correlation Analysis Of Stratum Deformation And Water Level Variation Of Aquifers And Land Subsidence Prediction Concerning Groundwater Exploitation And Recharge

The long-term exploitation of groundwater in Shanghai has resulted in the compression of aquifers as well as aquitards,which caused severe land subsidence.Since the 1960 s,Shanghai has adjusted its groundwater exploitation strategy by greatly reducing the pumping volumes and switching exploited aquifers from shallow layers to deep ones to control land subsidence.In this process,water level and stratum deformation have undergone significant changes,and the relationship between the variables is becoming increasingly complicated,which caused new spatial and temporal characteristics of land subsidence in recent years.In order to study the influence of groundwater exploitation and recharge as well as water level variation on stratum deformation in Shanghai,this paper analyzes the correlation between the variables using mathematical statistical methods based on long-term monitoring data.Afterwards,wavelet transform theory is adopted to study time-frequency characteristics of the variables,and the phase difference and lagging time are obtained.At last,land subsidence prediction model considering stratum deformation and lagging effect is established accordingly.The main contents are as follows:(1)Based on long-term monitoring data,the characteristics and correlations of groundwater exploitation and recharge,water level variation and stratum deformation are systematically analyzed.The results show that groundwater pumping volumes have greatly reduced,and the main exploited aquifers has switched from shallow layers to deep ones.The recharging volumes have been increasing in recent years,and the recharging proportion has developed from the second and third aquifers to the balanced distribution of each aquifer.It is obvious that controlling groundwater exploitation and increasing recharging volumes have significant effects on reducing stratum deformation.Correlations between variables are quantified by calculating Pearson correlation coefficient,and it can be found that there is a certain correlation between the variables,but the degree of linear correlation between stratum deformation and groundwater exploitation and recharge as well as water level variation is quite different.For aquifers,the second and the fourth aquifers have significant correlations both with groundwater exploitation and recharge as well as water level variation,while the third aquifer has a higher correlation with water level variation comparatively,which indicates groundwater exploitation and recharge have significant effect on aquifer deformation.For soft soil layers/aquitards,the degree of correlation with groundwater exploitation and recharge is higher,but the correlation with water level variation is relatively low.Considering the low permeability and slow water release of aquitards,the low correlation may suggest the existence of lagging effect.(2)The wavelet spectrums are calculated based on the wavelet theory,then the phase difference and lagging time between the stratum deformation and water level variation are obtained accordingly.The continuous wavelet transform is adopted to get the primary period between the variables and occurrence time.It is found that a distinct primary period of about 12 months generally exists between stratum deformation and water level variation,however the distribution time is different,which indicates the desynchrony between variables.Further research via cross-wavelet analysis found that lagging effect exists in each stratum,especially in aquitards,with lagging time reaching over one month.From the perspective of average lagging time,the lagging effect of shallow aquifers is greater than that of deep ones,while the aquitards show the opposite.Among them,the fifth soft soil layer has the longest lagging time of about 2.5 months.(3)Considering the two major factors of stratum deformation contribution and lagging effect,a polynomial distribution lagging model(PDL model)is established to predict land subsidence.Considering cumulative groundwater pumping volumes,cumulative deformation of the fourth aquifer and the fifth soft soil layer as variables,the lagging time and related parameters are determined,and the prediction model is established accordingly.The calculation results show that PDL model has great fitting effect with relative error less than 6%,which indicates its high accuracy and rationality to predict land subsidence in Shanghai.The prediction shows that under the condition of current groundwater exploitation pattern and steady annual stratum deformation,slight land lift will occur in the next 5 years,which verifies the controlling effect of restricting groundwater exploitation.