中国西南地区土地覆盖情景的时空模拟

气候植被类型的空间分布与土地覆盖类型的空间分布在时空层次上具有很好的相关性和一致性。在运用HLZ生态系统模型获得CMIP5的3种气候情景RCP26、RCP45、RCP85情景下西南地区未来90a(2011—2100年)HLZ生态系统时空分布情景数据的基础上,结合2010年土地覆盖现状数据,构建了土地覆盖情景的空间分析模型,并在此基础上,实现了西南地区未来90a土地覆盖情景的时空模拟分析。模拟结果表明:3种气候情景下,西南地区未来90a的落叶针叶林、落叶阔叶林、草地、耕地、冰雪、荒漠及裸岩石砾地等土地覆盖类型面积将呈逐渐减少趋势;常绿针叶林、常绿阔叶林、混交林、灌丛、湿地、建设用地、水体等土地覆盖类型面积则呈逐渐增加趋势。其中,湿地增加速度最快(平均每10a增加5.28%),荒漠及裸岩石砾地减少速度最快(平均每10a减少2.34%)。

Spatio-temporal simulation of land cover scenarios in southwestern of China

Numerous studies show that Karst area is more sensitive to influence of climate change and human activities, compared to the area in non-vulnerable condition. Therefore, quantitative simulating the land cover scenarios is important to understand the driving mechanism underlying land cover change in Karst area, and what policy should be carried out to prevent and reduce the land degradation. Especially, Karst area, as the typical ecological fragile zone, has been undertaking a series of ecological degradation, which have seriously affected the local socio-economic sustainable development. Karst area of Southwest China is one of the largest continuous Karst zone in the world, which major involves the Guizhou, Guangxi, Yunnan, Sichuan, and Chongqing province of China. This paper aims to develop a method for simulating the scenarios of land cover in Karst area and analyze its spatial distribution change under the global climate change. A simulation method of land cover scenario was developed on the basis of analyzing the correlation of spatial distribution between HLZ (Holdridge life zone) and the land cover, and the policy of basic farmland protection. According to the climate scenarios data of RCP26, RCP45, and RCP85 released by CMIP5 (the Fifth phase of the Coupled Model Intercomparison Project) and the land cover data in 2010 obtained from remoting sense images. Three land cover scenarios in Southwest China are respectively simulated in the next 90 years. The results show that three scenarios of land cover change have similar spatial landscape pattern and conversion trends. A gradual decrease was found in the following types of land cover, deciduous coniferous forest, deciduous broadleaf forest, grassland, cropland, ice and snow, and desert and bare rock; The other types of land cover would experience a moderate increase, namely, evergreen coniferous forest, evergreen broadleaf forests, mixed forest, scrublands, wetlands, construction land, water bodies and so on. Among the land cover types mentioned above, wetlands were projected to increase with the fastest rate (an increase of 5.28% per decade on average) and construction land were projected to increase most slowly (an average increase of 0.16% per decade), while desert and bare rock were forecasted to decline with the fastest rate (a decrease of 2.34% per decade on average) and cropland were forecasted to decrease most slowly (an average decrease of 0.26% per decade). It is worth noting that differences between land cover scenarios of 3 different climate scenarios lay in two aspects. On one hand, land cover scenario of every land cover type under RCP85 scenario stayed in top position in terms of the decadal change rate, especially, ice and snow decreasing far more than the other two scenarios. The next one was RCP45 scenario, land cover scenario of every land cover type under RCP26 scenario rank the last in terms of per decade change rate. On the other hand, each land cover type keep the same change trend under RCP85 and RCP45 scenarios during the following 90 years, while land cover scenario simulated with RCP26 data turn out to change in the opposite trend after 2070. Furthermore, the simulated result identify the method of land cover scenarios can avoid the difficulties which come from the complexity and uncertainty of mechanism analysis in land cover modeling, and suitable to simulate the land cover change on a regional scale.