The Evaluation And Prediction Of The Groundwater Quality In Jilin Area Based On Artificial Neural Network

Jilin city is located in Middle East of Jilin Province,which is 27120km 2 in area.The width from east to west is about 180 kilometers and the length from south to north is about 235 kilometers. The terrain is reduced from southeast to the Northwest gradually. Because of the earth tectonic movement of different periods, jilin area have formed the middle mountain area～the low mountain hills～the ravine lake district～river valley plain district. The climate belongs to the continental monsoon climate of the north temperate zone, makes a clear distinction between the four seasons in the city of Jilin. It's droughty and windy in spring, it is burning hot and rainy in summer, nice and cool and moist in autumn, cold and dry in winter, The over-years average rainfall is about 726.1 mm, the over-years water-surface evaporate is above 724.2 mm. The area lies in complex construction region, where the second land subsidence zone and the second raising zone join up .the groundwater widely distributed in jilin area is fissure water of clastic rock or pore water of clastic rock and fissure water of fluffy rock.With the increasing of the density of population of Jilin city and the development of national economy and the urban construction, the water requirement grows quickly, the waste water,emission of waste residue increase year by year, which aggravates imbalance between supply and demand of water resource and underground water pollution question . Therefor, it is significant for the local society and economic development to protect the groundwater source of water, according to the actual conditions to the groundwater water quality of this city, This paper deals with The evaluation and prediction of the Groundwater Quality in jilin area,the research results have important realistic meanings in utilizing and protection of the groundwater resources.Firstly, according to the whole country the second groundwater resources evaluation《The water quality investigation and a evaluation technological detailed rules and regulations》and basin characteristic and geological condition of Jilin area, confirm the evaluation and division principle of groundwater in Jilin and divide it. On this basis, analyzing the groundwater quality materials of the well in each division, divide and classify Jilin area groundwater chemical types based on 6 kinds of main ion (Na⁺, Ca²⁺, Mg²⁺, HCO₃^-, SO₄^2-, Cl^- -, K⁺ (merged in Na+) by C.A.舒卡列夫classification.According to The regulation of《underground water quality standard》GB/T14848-93, choose some Monitoring items such as SO42-, Cl-, M, CaCO3, NH4,Fe, NO3,etc. adopt the composite index means to evaluate the water quality of each divisions,meanwhile, design BP nerve network modelsoftware through vb software and set up the groundwater water quality of current situation and evaluate the model through the studying of the network (index year of current evaluation is 2000), evaluate the current situation to the water quality of every division. Next, do a comparative analysis of the evaluation result. thinking the grade result basically identical of two kinds of evaluation methods, therefore, the conclusion that the neural network model evaluation assessment gets is true and believable. Because of the mean of composite index have great influence on Fmax in specific and special case, whose water standard was confirmed according to more unfavorable thought, so the environmental quality that can still originally be accepted has been deteriorated. So the neural network model is more superior to it.The evaluation result indicates: The characteristic of water quality of groundwater of the whole district is shown as: the water quality of upper fengman basin, huifa river basin is better in the southeast, and in the northwestern part,â…¢water is distributed more centralized in lower fengman basin, lalin river basin,â… ,â…¡water is mainly distributed in the upper fengman huifa river basin, Jiaohe River basin, tuanshanzi River basin between the fengman reservoir and hada mountains reservoir, ersong master stream and wende river basin, lalin river basin and xilin river basin.â…¢water is distributed on the lalin river huolun river basin,kacha river basin, mangniu River basin between the fengman reservoir and hada mountains reservoir, ersong master stream and wende river basin, lalin river basin and xilin river basin.â…£water is maily distributed on the yinma river basin(excluding the yitong river),chalu river basin, olong River basin between the fengman reservoir and hada mountains reservoir.There are not too much groundwater materials of Jilin, so this paper evaluates the tendency of groundwater quality by two kinds of methods. Take the observation data of well 26101053 in the northwest of the soudengzhan town hospital of Jilin, choose four items : M, NH4, Cl-, Alkalinity from the observation date between 1998 and 2005 to set up model. tendency of groundwater. Firstly, Based on the observation data of groundwater quality , set up groundwater water quadratic exponential smoothing model, Smooth parameter adopts the following signal and chooses the method adaptively, which solved the smooth parameter perfectly,because of the complexity of the hydrogeological environment, the prediction result of the underground water pollution is not yet very ideal, therefor, have proposed the prediction model based on artificial neural network. The study samples of time arrays by sliding window to choose as input datas of the network, and survey value as the output data, the weighting of the input model is obtained through the network study, thus, the prediction model is established. Compared with the prediction errors of two kinds of prediction models mentioned above. this paper analyzes the reason that each model produces error, and discusses the advantage and deficiency of each model in predicting groundwater water quality, and adopt the neural network to predict four indexes of well 26101053 in the plan annual (2010) at last.According to analysis, The index of M, Cl-, Alkalinity of groundwater in the soudengzhan town of Jilin increase progressively, it is estimated that thecontent of M, Cl-, Alkalinity in the groundwater will reach 685.5 mg/1, 70.95 mg/1, 105.5mg/1 respectively in 2010, So it's necessary to protect the groundwater quality in this area to achieve the National Drinking Water Sanitation Standard.According to the groundwater water quality of the current evaluation and prediction of the Groundwater Quality in jilin area, It should take the following measures mainly in groundwater resources protection work of Jilin:1. Strengthen the prevention of the surface water pollution,control the pollution sources, Improve and rectify " the three wastes ". As to the sewage,we should improve the city drainage pipeline networks as soon as possible and strengthen the construction of facilities of sewage system.2.Develop green agriculture, popularize high-efficient, low toxicity, low residual pesticide chemical fertilizer actively, encourage developing green agriculture, lead and improve the water environment in the benign direction progressively.3.Forbid shallow layers of groundwater to pollute deep of groundwater, forbid the polluted shallow layers of groundwater and deep groundwater to mix and exploit, which should cause the shallow layer of groundwater to communicate with deep groundwater, water quality is obscured and pollute deep groundwater finally.4.Strengthen the dynamic monitoring of water quality of groundwater, evaluate and predict the groundwater water quality regularly.5. Strengthen people's thought and knowledge to the environment, arouse the whole people's action to protect the environment well.6. Active prevention is combined with scientific management, strengthen water environmental protection, make the rational water quality goal and corresponding local water environmental quality standards, make controlling the water resource have regulations to abide by.