Soil Quality Assessment And Influencing Factors In Suburb Of Taiyuan City

Soil is an important resource for human survival and a part of complex urban ecosystem.Suburb soil quality assessment and influence factor analysis has attracted much attention.Because of the different emphasis and evaluation standards traditional,soil quality assessment methods sometime can not reflect the real soil quality.In order to obtain more precise results of comprehensive soil quality assessment and explore its explanatory power of influence factors,we collected and analysed soil in April 2013,used posterior probability support vector machine model to evaluate the soil quality in Taiyuan.Then we used geographical detector model to discuss its explanatory power of main influence factors.The conclusions of this paper mainly concluded:(1)The contents of Pb,Cd and Hg in Taiyuan soil were mainly affected by human activities.It showed only a light level of soil metal pollution in research area.Except the mean concentration of Cr higher than the Chinese environmental quality standards.Among all the soil nutrients and heavy metals existed significantly positive relationships.(2)The main factors which influencing the content and distribution of each soil element as follows:industrial and vehicular emissions;represented sewage irrigation;phosphorus and potassium compound fertilizer application;represented electroplating factory pollution.Pd was controlled by industrial pollution,vehicular emission and sewage irrigation.(3)we used the posterior probability SVM to assess soil quality and got the category and probability information of samples.The 55%of soil samples in Yingze district and Jinyuan district belonged to primary standard.The 58.06%of soil samplesbelonged to secondary standard in north of Taiyuan city.The 52.63%of soil samples belonged to secondary standard in southwest of Taiyuan and 70.27%of soil samples belonged to secondary standard in southeast of Taiyuan.(4)The mainly influence factors of primary standard soil samples were DEM,soil types and land use types.The mainly influence factors of third standard soil samples were NDVI,population and land use types.Any two factor interactions on the soil quality were stronger than the influence of single factor.(5)The posterior probability of support vector machine model could get more accurate qualitative and quantitative sample classification information.Evaluation should be paid attention to formulate reasonable evaluation standard.In order to got more comprehensive diagnosis,we used geographical probe model which should provide more effective soil physical and chemical factors,meteorological factors and human factors.The results of this study had theoretical verified the applicability and superiority of posterior probability in the ecological evaluations and made up for a lack of the some traditional evaluation methods.The comprehensive soil quality assessment model has clear structure and easy operation also suit for other environmental assessments such as atmosphere,water quality assessment.Geographical detector model has obvious superiority on diagnosing the influence factors about spatial distribution of geographic things.Using this model,geographical factors can draw the explanatory power of geographical things' spatial distribution.The above methods provide certain reference for pollution control and environmental protection implemented by the relevant departments.