Research On Real Time Groundwater Quality Prediction Model Of Lower Liaohe Plain

The thesis is supported by two subjects of National Key Technology R&D Program in the 11th five-year Plan which are "Groundwater numerical model of central plain in Liaoning Province" (No.2007BAB28B04-03) and "Real time groundwater prediction model of central urban agglomeration in Liaoning Province" (No.2006BAB04A09-02) and one subject of Doctoral Fund of Education Ministry which is "Release rule of fluorine in water and soil of fluorine highly concentrated lake in diversion project from Nenjiang to Baicheng City" (No.200801830044). The central urban agglomeration of Liaoning Province located in Lower Liaohe Plain is the economic centre of northeast Asia and also an important grain producing area, using huge quantity of industrial water and agricultural water. Groundwater quality is gradually getting worse under human activities. How to predict groundwater quality properly and fast is very significant to guarantee groundwater development in Lower Liaohe Plain. Geo-statistics method was used to analyze the distribution of Quaternary Aquifer. Groundwater flow numerical model was built with numerical simulation technique. The main types of hydro-geochemical processes along groundwater flow path were determined by inverse hydro-geochemical simulation. Groundwater solute transport model considering of main hydro-geochemical processes were built on the basis of groundwater flow model. The two models were coupled by computer programming, and the function of collecting and dealing with groundwater real time monitoring data was also developed. Finally, the real time groundwater quality prediction system under operation interface was established.In order to determine the distribution of Quaternary aquifers and whether stable clayey existing in a wide range of the study area, the lithology are generalized into four types of clayey, silt clay, sand and gravels, and geological structure model was established with T-Progs, which is a transition probability geo-statistic software. The study area was meshed into 126 rows,80 columns 94 layers, and 536458 grids in total. The length and width of horizontal grids were both 2000 m, and the maximum length was 5 m in vertical. The transition probability of different lithology along vertical, strike and dip direction was calculated, and the geological structure model was established finally. From joint geological cross sections and probability distribution chart, it is shown that there is only small range of clayey layers existing in Quaternary aquifers. The clayey layers can not form an aquitard; therefore, the whole Quaternary aquifers can be generalized into one large aquifer.After determination of aquifer distribution, the groundwater flow characteristics and boundary conditions were generalized, and the conceptual hydro-geological model was established. The mathematical model of heterogeneous isotropic Quaternary aquifer and transient flow was built. The model was calculated with groundwater modeling code MODFLOW 2000, which is also the kernel program of GMS. The study area was meshed into 230 columns,221 rows, and 30808 grids in total. The length and width of each grid were both 1000 m, with the area of 1 km2. Seven years groundwater data from January 2000 to January 2007 were used to calibrate and validate the model with setting the stress period of one month and time step of 15 days. The model was run after inputting all sources and sinks data into the model. The calculated and actual measured groundwater levels were fitted. It is shown from the regional and single well fitting curve of groundwater level that the calculated and actual measured levels were fitted well. The nodes with fitting errors lower than 0.5 m took up more than 80% of all known nodes. The model could truly reflect the actual features of Quaternary aquifer in Lower Liaohe Plain, and the calibrated parameters were reliable. The groundwater flow numerical model provided the hydro dynamic model basis for real time groundwater quality prediction model.The MT3DMS code would be used to build groundwater solute transport model, which can consider of advection-dispersion, mix and evaporation concentration processes, but can not describe the water-rock interaction along groundwater flow path. Inverse hydro-geochemical simulation method was used to determine the mineral dissolution/precipitation and cation exchange processes. The dissolution and precipitation of calcite, dolomite and gypsum occurred all over the study area. The main process of halite is dissolution, and the precipitation of halite only occurred near saline water area. The cation exchange process mainly occurred in the middle and lower of flow path. The processes of mineral dissolution/precipitation and cation exchange made little contribution to variation of groundwater compositions from the time scale of several years to decades and the macro spatial area of whole Lower Liaohe Plain. Therefore, the advection-dispersion, mix and evaporation concentration processes were mainly considered during building groundwater solute transport model.The chloridion and TDS were selected as simulation factors. The chloridion and TDS carried by rainfall, river and irrigation water were treated as external sources. Also, the impact on ion concentration by evaporation concentration was considered. The groundwater solute transport model of Lower Liaohe Plain was built and calculated by kernel code MT3DMS of GMS. The calculation of transport model was based on groundwater flow model; therefore, the spatial and temporal discretization was set as the same as flow model. The model was calibrated and validated according to actual measured data. The calculated and measured concentration contours were fitted well, indicating the high precision and reliability of the model. Groundwater solute transport model and groundwater flow model of Lower Liaohe Plain were the basis of real time groundwater quality prediction model.The MF2K and MT3DMS code were partially modified and compiled according to actual demands on the basis of fully mastering their structures. The two models were used as the kernel programs of real time groundwater quality prediction model. The calibrated and validated groundwater flow model and groundwater solute transport model were the basis of the final prediction model. The VB language was used to program a code to link and couple the two models, and the interfaces between these two models and the real time groundwater monitoring system were developed separately, achieving the function of collecting and dealing with real time data. The real time groundwater quality prediction system was finally established with the integration of groundwater flow model, groundwater solute transport model and the interfaces to real time groundwater monitoring system. The system can collect real time data of groundwater level and quality, and predict future groundwater level and groundwater solute concentration immediately, providing technique support for scientific decision of groundwater resources managers.