Study on driving forces of wetland change in the Western Liaohe River basin based on random forest model

International Symposium on Resource Exploration and Environmental Science

地球科学;生态环境科学

Wu, Menghong^1 ; Yang, Changbao^1 ; Zhang, Yanhong^1 ; Lin, Nan^2

School of College of Geo-exploratin Science and Technology, Jilin University, Changchun, China^1

School of College of Surveying and Prospecting Engineering, Jilin Jianzhu University, Changchun, China^2

关键词: Driving factors;    Human activities;    Natural factors;    Partial correlation;    Population densities;    Prediction accuracy;    Random forest modeling;    Temperature factor;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/64/1/012009/pdf
DOI  :  10.1088/1755-1315/64/1/012009

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

Based on the platform of RS and GIS, random forest progression model is used for study driving force of wetland change in western Liaohe river basin, five influencing factors which include elevation, slope, temperature, precipitation and population density are chosen to establish random forest progression model about the wetland change and the driving factors. Using the the mean value of the prediction accuracy outside the bag calculated by the model to evaluate the importance of the variables. The result indicates that the coefficient of partial correlation between precipitation and wetland density is the largest among the five influencing factors, followed by temperature, population density, elevation and slope is smallest. The influence of natural factors on the change of wetland density is mainly reflected in precipitation and temperature factors, and the precipitation is obviously higher than that of temperature, under the influence of human factors, the influence of population density factor on wetland density is higher than that of elevation and slope factor. The result shows that in the past 40 years, the human activities in the study area have increased the density of wetland to some extent, but it is not the main factor.