地形
生态学
草原
地理
自然景观
环境科学
自然地理学
环境资源管理
自然(考古学)
生物
考古
作者
Lianqing Xue,Boli Zhu,Yiping Wu,Guanghui Wei,Shumin Liao,Changbing Yang,Jing Wang,Hui Zhang,Lei Ren,Qiang Han
标识
DOI:10.1016/j.scitotenv.2018.10.382
摘要
With large-scale developments, the Manas River Basin (MRB) is in an extreme imbalance especially in land use, thus causing a series of ecological problems. A reliable dynamic ecological risk assessment is expected to provide useful information for the economic development. Through coupling spatial Cellular Automaton-Markov (CA-Markov) model and Landsat satellite images in 2000, 2008 and 2016, we forecasted the land use maps in 2024 and 2032. Based on the ecological risk model, we evaluated the ecological risk at landscape level from 2000 to 2032. More importantly, an improved evaluation of ecological risk was proposed based on terrain gradients and the correlation between terrain niche index (TNI) and future ecological risk was analyzed. The results showed that the artificial oases and urban are expanding, while the natural grassland is shrinking. Corresponding to the rapid development stage and stable consolidation stage, farmland will be followed by a slower increase (2016–2032) after a rapid increase (2000–2016), and water decreases first but then is projected to recover. As the overall spatial diversity increasing, the ecological risk in the whole basin is growing, especially in grassland. Compared with the stable critical state in artificial landscape, the future ecological risks in natural landscape tend to increase due to the cumulative effects of human activities. Also, we found that the great ecological risk mainly happens in “high altitude and complex terrain” or “low altitude and flat terrain” areas. The future ecological risk in medium terrain niche index (TNI) gradient will increase, while it will decrease in the lowest. Above all, the proposed framework can do well in forecasting ecological risk at landscape level, and can help simply infer the changes of ecological risk based on terrain.
科研通智能强力驱动
Strongly Powered by AbleSci AI