The Research Of Groundwater Flow Modeling In Ningxia Zhongwei Plain With FEFLOW Model

The study area is located on Zhongwei plain in the Ningxia Hui Autonomous Region, it is dry with strong evaporation. The Yellow river is the south boundary in the study area, which supplies the convenient use of water resources, but perennial irrigation causes groundwater level to rise and soil salinization, and also, the variegated irrigation canals and drainage ditches disturb groundwater flow. This research analyzed all the factors influencing the groundwater dynamic, such as the study area boundaries, the position of irrigation canals and drainage ditches, landuse, precipitation recharge, groundwater evaporation, the number and the position of production wells, and then calculated the volume of recharge and drainage for groundwater. According to data requirement from FEFLOW, ArcGIS is used to process the data into vector layers. After inputting data, setting and adjusting parameters through FEFLOW model, the groundwater flow is simulated under steady and unsteady conditions.Since the shortage of initial hydraulic head in unsteady condition in2010, the steady simulation results must be used to unsteady simulation. After a series of operations and analyses including building steady model, adjusting the parameters many times, and then analyzing the distribution of simulating groundwater flow, groundwater levels in monitoring sites, the error accuracy and groundwater balance, the simulation results are good and parameters calibration is reasonable, therefore, it is reliable to make the steady simulation results as the initial hydraulic head in unsteady condition.In unsteady condition, the data in source and sink parameter is assigned to time series by seasonal irrigation and drainage, and meteorological conditions. The time unite is one month. The model is adjusted to unsteady condition and the simulation period is365days. Adjusting the source and sink many times, then the water level differences in six monitoring wells are in2meters between the simulation results and monitoring results, and the variation trends are also roughly consistent, which further the validation of the simulation accuracy.However, there are some differences between the simulation results and the monitoring results, which are caused by many uncertain factors. This paper analyzed the uncertainty from the input of data, model structures and model parameters. These factors can be summarized as the following points:the error in data monitoring, the representativeness of monitoring sites, the bias of coefficient selection in indirect calculation, the simplification of irrigation canals and drainage ditches, the hypothesis and simplification of nature in the model, the scale of parameters in the model, and so on.After calibrating parameters and validating the good fitting degree, then the sensitivity of input water is analyzed. Sensitivity analyses choose four factors including the leakage of irrigation canals, the discharge of drainage ditches, boundaries fluid and conductivity, which are changed by±10%and±20%to research the change of groundwater level. Comparing the influencing degree of four parameters by absolute average, the sensitivity in turn is the boundary fluid, the discharge of drainage ditches, the leakage of irrigation canals, conductivity. With that result, the discharge of drainage ditches is the workable and effective method to control the groundwater level in study area.