China is undergoing rapid urbanization, which has brought great pressure on croplands throughout the country, especially in fast developing cities, such as Hangzhou. In this study, an attempt was made to monitor and model the cropland dynamics of Hangzhou from 1990 to 2035. The spatial-temporal changes in the cropland were discussed based on the land cover maps along with urban-rural gradient analysis. After understanding the spatial-temporal patterns of cropland changes, the cellular automata-Markov model was employed using the historical land cover maps and other explanatory data to perform a scenario-based simulation. Accordingly, three scenarios, namely spontaneous scenario (SS), protected area ensuring scenario (PAES), and optimal agriculture developing scenario (OADS), were designed for simulating the cropland distribution in 2035. The monitoring results showed that during 1990-2015, the cropland area decreased 1512.46km2 under rapid urbanization. Areas at a distance of 12km from the city center experienced maximum cropland loss. Among all the spatial metrics, aggregation index of the cropland exhibited the highest correlation with the distance to the city center (r=0.77 in 2015), thereby suggesting an obvious trend in aggregation along the urban-rural gradient. The modelling results reported that under PAES and OADS, the study area could gain 81.76km2 and 255.14km2 more cropland, respectively, than that under SS in 2035. Thus, policies applied in PAES and OADS would be effective for cropland protection.
Keywords: CA-Markov model; Land architecture metrics; Scenario-based modelling; Urban-rural gradient; Urbanization.
Copyright © 2019 Elsevier B.V. All rights reserved.