Geo-environmental suitability assessment for agricultural land in the rural–urban fringe using BPNN and GIS: a case study of Hangzhou

Abstract

It is the hotspot issue of the land use and land cover change in assessing the suitability for agricultural land in the rural–urban fringe. This paper presents an integrated technique using back-propagation neural network (BPNN) and geographic information system (GIS) to assess suitability for agricultural land based on geo-environmental factors in the rural–urban fringe. Hangzhou was chosen for the case study. Four groups comprising ten separate sub-factors of geo-environmental attributes were selected as major factors affecting agricultural land. A back-propagation algorithm was used to calculate the weights by adjusting the number of hidden nodes and the learning rate. A suitability assessment model was established based on the above sub-factors with their corresponding weights, along with field survey data, BPNN, and GIS technology. According to this model, the geo-environmental suitability for agricultural land in the rural–urban fringe of Hangzhou was divided into four levels. The numerical evaluation results demonstrated that 52.95 % of areas in the rural–urban fringe in Hangzhou are suitable for agricultural land and that the current land use is appropriate. Furthermore, the results proved that applying BPNN and GIS is a very effective method for assessing the agricultural land suitability based on geo-environmental factors. The research results could provide support for agricultural land planning and management, to promote the sustainable use and protection of land resources in the rural–urban fringe in Hangzhou.