#L07 Global LCLUC drivers

Abstract

Uncovering the historical land cover and land use change (LCLUC) drivers is vital for projecting the future LCLUC. Due to the differences from spatiotemporal scales, methods and LCLUC datasets, previous studies have not reached a consensus regarding the global LCLUC drivers in terms of biophysical and socioeconomic aspects. Here we present a global scale estimation of the biophysical and socioeconomic drivers of several dominant LCLUC types, e.g. deforestation, agriculture intensification, and urbanization over the past century (1900~2015). The LCULC drivers are extremely complicated with great variances in both spatial and temporal scales. More advanced models are needed to account for the spatial and temporal heterogeneities. We first use a time-series segmentation technique to identify different periods for each LCLUC type, and analyze the key LCLUC features within each period globally. Then we detect the hotspot regions in term of Agro Ecological Zones (AEZs) for each LCLUC type. Finally, we employ the geographically weighted logistic regression to assess the spatially explicit drivers of LCLUC in each period.

The spatial and temporal patterns of major LCLUC types revealed by our study can provide a perspective in the overall global LCLUC characters at a centennial time scale. The spatially and temporally explicit LCLUC drivers also promote the understanding of the complex LCLUC processes. The identified biophysical and socioeconomic components will be coupled with an ACME to improve the accuracy of future projections of LCLUC. The methodology developed in this study will also serve as a comprehensive example for LCLUC studies at different spatiotemporal scales.