Assessment of Grassland Degradation and its Mechanisms using VHR Imagery

Grasslands are ecologically heterogeneous ecosystems in an often environmental and socioeconomic marginal position. Their unsustainable land management leads to land degradation, causing desertification and negatively affecting the environment and economy at a global level. This is the case in Inner Mongolia, China. Therefore, monitoring land degradation and understanding desertification mechanisms are important for sustaining policymaking and sustainable land management. This study aims to advance desertification conceptualisation and monitoring with remote sensing. Remote sensing is the most used and suitable technique to study desertification. However, the use of low and medium spatial resolution hinders comprehensive assessments. This research investigates the potential of using VHR images to monitor grassland degradation in Hexigten Banner, Inner Mongolia. The study is based on a visual interpretation analysis of the open-access Google Earth imagery. The desertification attributes are recognised and grassland degradation is assessed as a combination of vegetation and soil degradation. The analysis shows a large extent of degradation in fenced and unfenced grasslands and higher degradation severities close to settlements, croplands, and water points. A conceptual framework of desertification processes is developed and used to understand and explain the land degradation patterns in the study area. The model considers the cross-scalar mechanisms between the ecological, policy, social, and economic domains in a coupled human-environment system. In this study case, it can be inferred that grassland degradation is the consequence of cropland and built-up land expansion and displacement, which have reduced and marginalised the rangelands, exploitation of natural resources close to settlements and water sources, and overgrazing in fenced and unfenced grasslands caused by land fragmentation and socioeconomic imbalances. In conclusion, the VHR image analysis and a cross-scalar multidomain conceptual framework revealed to be tools that can help policy and land management decision making. Therefore, further improvements are recommended.