Monitoring project impact on biomass increase in the context of the Great Green Wall for the Sahara and Sahel Initiative in Senegal

摘要

Land degradation and desertification represent a major threat to the population and ecosystems of (semi)-arid regions like the Sahel and the Sahara. In 2007, the African Union launched a pan-African programme, the Great Green Wall for the Sahara and the Sahel Initiative (GGWSSI) to reverse land degradation and desertification in the region, improve food security and support local people to adapt to climate change. Within the GGWSSI different kinds of projects have been implemented. In order to quickly evaluate the effectiveness of (a large number of) restoration and sustainable rangeland/agriculture projects, a methodology to remotely assess the biophysical impacts of diverse interventions is desirable. Within the Administrative Arrangement Technical and scientific Support to agriculture and Food and Nutrition Security we aim to develop a multi-scale remote sensing (RS) based approach to monitor the biophysical impact of sustainable agriculture and rangeland projects. Specifically, we propose to utilize a biophysical indicator obtained via satellite imagery and compare the difference between project sites and corresponding reference sites before and after the intervention. Besides this specific comparison between project and reference sites, the general state and development of vegetation and other parameters (like precipitation) on a larger scale are important information for project planning in general and the project impact assessment. This report focuses on this general analysis of vegetation and precipitation trends in north Senegal and provides some first basic approaches for RS-based impact assessment on biomass increase of selected GGWSSI projects. An upcoming paper will concentrate on a more advanced approach of impact assessment. Time series of satellite-derived precipitation estimates (P, the main driver of vegetation growth in the area) and Normalized Difference Vegetation Index (NDVI, indicating vegetation amount and health status) data were utilised to characterize general precipitation and vegetation characteristics in the region and to compute long- (for P) and short-term linear trends (for P and NDVI). This is important to distinguish between general climatic trends in the region and vegetation trends due to project intervention. The results indicate a significant long-term increase in annual precipitation sums over the period 1981–2014 in the study area, while there is no significant precipitation trend in the more recent (and shorter) time period (2001–2014). The NDVI-based analysis of vegetation revealed some local positive and negative trends. As there is no significant precipitation trend over this time, we assume that other/additional factors than precipitation changes need to be considered as drivers for the vegetation trends. For an in-depth analysis of local vegetation changes (also with respect to land degradation) future studies should also include other sources of information (e.g. field studies, interviews of local people).The basic assessment of possible biophysical impacts of selected restoration projects was done via visual inspection of very high resolution images before and after the intervention and by computing differences of maxNDVI between project sites and reference sites. For some projects, slight positive changes after the intervention could be observed, indicating an increase of biomass. Other projects did not show any visible positive effect of the interventions. Further research is currently in progress to develop a more automatized and statistically sound method for this type of comparison. Should this advanced approach confirm the failure of some projects, further studies should be undertaken to understand the reasons for the failure to guide future interventions and project planning/monitoring.