Dynamic Analysis Of Groundwater And The Prediction Of Groundwater Level Change In One Plain Area

The research area is located in Chengdu,Sichuan.The large-scale groundwater exploitation in the area leads to a drop in the groundwater level and predicts the groundwater dynamics in the entire plain area.Geological and hydrogeological surveys were conducted in the study area,water level of the High water period,flat water period and low water period and water quality measurement data of 161people’s wells in the study area were analyzed,and the dynamic types of groundwater were analyzed;mathematical statistical characteristics analysis and GIS technology were used to focus on the analysis of groundwater Distribution characteristics,discussing the annual variation characteristics of water level burial depth;at the same time,the water chemical distribution characteristics were analyzed,and the BP neural network was used to predict the water level burial depth of the study area;and the flow after 5 years and 10 years under the two conditions The field was predicted.The following main results were obtained:（1）The dynamic types of groundwater in the plain area of Chengdu can be divided into four types:rainfall-evaporation type,agricultural irrigation-rainfall type,river-rainfall type,and human activity influence type.（2）The burial depth of the water level in the study period is 2-4m,and the burial depth of the local water level is less than 2m.In some areas,the burial depth is 4-6m.The overall distribution law of the High water period,flat water period and low water period burial depth is basically the same.The high water period is generally 0-1.5m higher than the normal water period,and the local water level is more than 2m.The maximum can reach 3.24m.（3）Analyzed the water chemical composition data of 49 groups of groundwater samples in the study area.The results showed that the water chemistry type in the study area was HCO3-Ca;the main components were spatially interpolated and their distribution characteristics were analyzed,such as areas with high total hardness and the following Fu gypsum is related to the glauber’s salt layer;and the correlation of its main components is analyzed.For example,TDS is highly linearly related to Na⁺,Ca²⁺,Mg²⁺,and SO₄²⁺ions.The PH values of the three periods of Feng-Ping-Ku in the study area are all between 6.4-8.7,and the overall distribution law and regional change trend of High water period,flat water period and low water period are basically consistent with that of the Pingshui period.The three phases of High water period,flat water period and low water period in the study area met the II and III categories of groundwater quality classification standards.The TDS concentration in the high water period was lower than that in the normal water period,and the decrease range was between 0-150 mg/l.But its overall distribution law is basically consistent with the flat water period.（4）The correlation between groundwater level and temperature,humidity,evaporation,and precipitation is analyzed.Using these four parameters as the input values of BP neural network,a BP neural network simulation model is established.The groundwater real value and the predicted value fit well.Using the built model to predict groundwater burial depth,the prediction results of groundwater burial depth from August 2019 to July 2020 show that the future groundwater burial depth is not much different between the high water period and the low water period,the maximum and minimum burial depths are 2.96m and 2.22m,the change is 0.74m.（5）The hydrogeological conditions in the study area were generalized,and a three-dimensional hydrogeological conceptual model was established.After the three-dimensional unsteady flow numerical simulation model in the study area has been identified and tested,the simulated water level of the typical observation well and the measured water level have a good fitting effect.For the two groundwater mining schemes,the groundwater level changes after 5 years and 10 years under the two schemes are predicted.The water level of scheme 1 generally rises with an average increase of 1.4m,and the water level of scheme 2 declines overall,with an average decrease of 1.68m.The water level predicted by the numerical simulation is compared with the prediction result of the BP neural network.The buried depth of the water level is not much different.