Impact of Anthropogenic Land-uses on Salinization in the Yellow River Delta, China: Using a New Rs-gis Statistical Model

Land salinization is threatened with secondary salinization recently because of unreasonable exploitation of natural resources and global change. However, previous assessments cannot obtain comprehensive and spatial explicit statistics and impede to develop ecologically sound policies for government. This study tries to confront these difficulties and identify key anthropogenic determinants, their extent and effectiveness (positive or negative) in the YRD. A SAR model was firstly applied to evaluate human impacts on salinity. Based on RS-GIS system, three sub-region models (LF, ST and LU) were created to normalize salinity heterogeneity. The predicted land salinity was served as the response variable in models. 12 explainable variables were elaborately selected as input into the models. The conventional OLSR model was compared simultaneously to validate performance of SAR model. SAR model fitted better than OLSR model because spatial autocorrelation in soil salinity was well dealt with. SAR modelling in sub-regions where environmental impacts were normalized had a high accuracy in determining the most influenced factors. Among established key determinants, oil exploitation, road construction and saline aquaculture were aggregative to salinization but only in originally highly saline sub-regions, such as LF_coastal zone and ST_soil III. Two agricultural activities, salt-tolerant crop plantation and fertilization, were ameliorative to salinization in most models. The most effective agricultural alleviation occurred in moderately saline sub-regions, such as LF_floodplain, ST_soil II and LU_wastland to farmland, which may benefit by the development of farm forests and farm ponds. The assessment with SAR sub-region model provides suggestions for policy makers and also has implications in other environmental assessments.