Mapping gains and losses in woody vegetation across global tropical drylands

Woody vegetation in global tropical drylands is of significant importance for both the inter-annual variability of the carbon cycle and local livelihoods. Satellite observations over the past decades provide a unique way to assess the vegetation long-term dynamics across biomes worldwide. Yet, the actual changes in the woody vegetation are always hidden by inter-annual fluctuations of the leaf density, because the most widely used remote sensing data are primarily related to the photosynthetically active vegetation components. Here, we quantify the temporal trends of the non-photosynthetic woody components (i.e. stems and branches) in global tropical drylands during 2000–2012 using the vegetation optical depth (VOD), retrieved from passive microwave observations. This is achieved by a novel method focusing on the dry season period to minimize the influence of herbaceous vegetation, and using MODIS (MODerate resolution Imaging Spectroradiometer) NDVI (Normalized Difference Vegetation Index) data to remove the inter-annual fluctuation of the woody leaf component. We revealed significant trends (p < 0.05) in the woody component (VODwood) in 35% of the areas characterized by a non-significant trend in the leaf component (VODleaf modeled from NDVI), indicating pronounced gradual growth/decline in woody vegetation not captured by traditional assessments. The method is validated using a unique record of ground measurements from the semi-arid Sahel and shows a strong agreement between changes in VODwood and changes in ground observed woody cover (r2 = 0.78). Reliability of the obtained woody component trends is also supported by a review of relevant literatures for eight hot-spot regions of change. The proposed approach is expected to contribute to an improved assessment of e.g. changes in dryland carbon pools. This article is protected by copyright. All rights reserved.