An integrated simulation framework for NDVI pattern variations with dual society-nature drives: A case study in Baiyangdian Wetland, North China

Due to climate change and socio-economic development, the growth and distribution of wetland vegetation have been destroyed, resulting in a degradation trend of wetland ecosystem. It is critical to simulate variations in the normalized difference vegetation index (NDVI) patterns under changing environments to guide targeted restoration and protection. Considering that current NDVI pattern simulation focused only on the linear driving forces of nature or society and had a low simulation efficiency, an integrated simulation framework with dual societynature drives was proposed. In the framework, the spatial and NDVI patterns were simulated by system dynamics-future land use simulation (SD-FLUS) and random forest (RF) models, respectively. Baiyangdian Wetland in North China was taken as a case study, and its typical vegetation-reed-was taken as the simulation object. The NDVI patterns were simulated under the effects of socio-economic development, climate change and ecological protection policies. The results showed that the spatial patterns were affected by social factors, while the NDVI patterns were affected by natural factors. The spatial aggregations of the NDVI were affected by social and natural factors. Besides, the growth space of reeds was easily encroached under the existing development and protection policies. If ecological protection policies were implemented, the reed area would be restored, but the growth status would be poorer, requiring further protection measures and more attention to the growth status of reeds in sensitive areas. This study provides a new research perspective for simulating NDVI patterns and scientific supports for refined management of wetland vegetation.