Land use and landscape pattern changes of Weihai, China based on object-oriented SVM classification from Landsat MSS/TM/OLI images

Abstract

Weihai's urban development model is representative of coastal cities in China. Landsat MSS/TM/OLI images were used to extract the land use types of Weihai from 1985–2015 using the object-oriented support vector machine (SVM) classification method. The landscape pattern indexes were calculated based on the classification results of land use. The temporal and spatial characteristics of land use and landscape pattern were analyzed by considering Weihai’s economic development. The overall kappa (OK) coefficients were greater than 83% and the overall accuracy (OA) values were greater than 85%. Over 80% of both producer’s accuracy (PA) and user’s accuracy (UA) values of different land use types were greater than 85%. Changes in land use and landscape patterns were closely related to the local economic development. The areas of cultivated land and woodland continued to decrease. The expansion of construction land came mainly from the cultivated land and woodland as with as very little from waters. Under the combined effect of economic development and human activities, the landscape pattern of Weihai as a whole has been decentralized. The internal structure is relatively complex and the fragmentation of land use is increasing. The development model of Weihai is the expansion-aggregation-expansion.