Landscape Pattern Analysis and Prediction in Qingyang City Based on CA-Markov Model

In this paper, the spatial-temporal landscape pattern changes of Qingyang city were analyzed based on landscape pattern indices from the classified remotely sensed data of 2000, 2006 and 2012. The future landscape pattern in 2018 was predicted using the CA-Markov model. The results suggested that cropland, forestland, and grassland dominated the landscape pattern in Qingyang city, with a total percentage of more than 90%, and the area of unused land, construction land, and water body area successively decreased. During the period from 2000 to 2012, the percentage of cropland kept on reduced while those of other landscape types showed an upward trend, especially the grassland with the highest increase rate because of the policy of returning cropland back to forestland and grassland. Moreover, the patch numbers, as well as the fragmentation degree of cropland, forestland, and grassland were larger than other landscape types. During the period from 2000 to 2012, the patch density of cropland, forestland, and construction land presented a growing trend, and the landscape of cropland, forestland and grassland were spottily distributed. The natural connectivity of blocks was relatively high, and the degree of patch aggregation in grassland and construction land increased, of which the construction land had more changes. The prediction results based on the CA-Markov model indicated that the cropland would continue to decrease in 2018, and the other five landscape types would keep on increasing in Qingyang city. Under this trend, the cropland, forestland, and grassland are still the dominant types in this study area. This study can contribute to decision-making on ecological environmental protection and land use planning in Qingyang City.