Temporal Analysis And Evaluation Of Groundwater Quality In The Central Area Of Fuzhou Basin

Lack of research of shallow groundwater in Fuzhou basin,it is important to study the temporal and spatial variation of shallow groundwater quality in the area,to reasonable exploit protect and utilize the groundwater.Based on the monitoring data of shallow groundwater in the central area of Fuzhou basin from January to July 2010?2014,the general of statistics,spatial analysis,SVM and PSO,the water quality spatial and temporal variation characteristics and the influence factors of water quality single index is analyzed,the water quality spatial interpolation model based on neural network is studied,and the PSO-SVM method of water quality assessment is studied.The conclusions of this study are as follows.1.This thesis takes statistical method and spatial analysis technology to study central area of Fuzhou Basin's shallow groundwater spatial and temporal variation characteristic.Statistical analysis shows that the nitrate,ammonia nitrogen,fluoride and iron in the study area were polluted in different degree,and the pollution of nitrate and fluoride was serious.The content of total hardness,total dissolved solids and nitrate show a trend of deterioration,the Jin'an district's south area shallow groundwater quality index show an obvious degradation trend.As the first principal factor,total hardness influenced by geologic lithology and soil type,total dissolved solids and fluoride as the second principal factor influenced by precipitation,evaporation capacity and grade,nitrate,ammonia nitrogen and iron as the third factor influenced by human activities.2.Combining given points and influence factors to build a groun.dwater quality spatial interpolation model based on RBF neural network.Extracting spatial distance between given points and interpolation points,water quality index as groundwater quality spatial interpolation model based on RBF neural network's input factors,the model simulates water quality indexes' spatial variation and use cross-validation to find the method receive a good simulation precision and high efficiency.Total hardness is main I and II level.Total dissolved solids is main II and III level,and the groundwater quality is poor in Cangshan district,Gulou district and its surrounding areas.Total hardness and total dissolved solids show a serious degradation trend even if they haven't be polluted.Fluoride content in the junction of Jin'an district,Gulou district and Cangshan district is higher,it has pollution phenomenon,and it shows an optimization tendency.The south area of Fuzhou has a high human activity strength,the nitrate content is exceeding,and becomes worse since 2011.Ammonia nitrogen and iron have a good state and become better,iron show a bad distribution in the central area.3.Building a water quality assessment model based on PSO-SVM method assess groundwater quality,to analyze groundwater quality spatial and temporal variation in the central area of Fuzhou Basin.Experimental result demonstrates that standard SVM is sensitive to predisposing factor,BP neural network has an instability,and the result of PSO-SVM is rather in agreement with the practical condition.The integrated groundwater quality of central area of Fuzhou Basin is general,main are ?,? level straggling distributed in study area and its proportion reduced dramatically.Gulou district and its surrounding districts water quality are worse,and these districts are mainly distribution area of ? level of groundwater quality.groundwater quality take on degradation trend,it mainly changes from ? level to ? level.