Global‐scale mapping of changes in ecosystem functioning from earth observation‐based trends in total and recurrent vegetation

Abstract

Aim To evaluate trend analysis of earth observation ( EO ) dense time series as a new way of describing and mapping changes in ecosystem functioning at regional to global scales. Spatio‐temporal patterns of change covering 1982–2011 are discussed in the context of changes in land use and land cover ( LULCC ). Location Global. Methods This study takes advantage of the different phenological cycles of recurrent vegetation (herbaceous vegetation) and persistent vegetation (woody/shrub cover) in combining trend analyses of global‐scale vegetation based on different annual/seasonal normalized difference vegetation index ( NDVI ) metrics. Spatial patterns of combined vegetation trends derived from the G lobal I nventory M odeling and M apping S tudies NDVI are analysed using land‐cover information ( GLC 2000). Results The direction of change in annual and seasonal NDVI metrics is similar for most global terrestrial ecosystems, but areas of diverging trends were also observed for certain regions across the globe. These areas are shown to be dominated by land‐cover classes of deciduous forest in tropical/subtropical areas. Areas of observed change are found in dry deciduous forest in South America and central southern Africa and are in accordance with studies of hotspot LULCC areas conducted at local and regional scales. The results show that dense time series of EO data can be used to map large‐scale changes in ecosystem functional type that are due to forest cover dynamics, including forest degradation, deforestation/reforestation and bush encroachment. Main conclusions We show that areas characterized by changes in ecosystem functioning governed by LULCC at regional and global scales can be mapped from dense time series of global EO data. The patterns of diverging NDVI metric trends can be used as a reference in evaluating the impacts of environmental changes related to LULCC and the approach may be used to detect changes in ecosystem functioning over time.

Bibliographic Information