Sustainable Use of African Wetlands for Food Security: A Spatial Evaluation Approach
Stefanie Steinbach is a PhD student in the Department of Natural Resources. (Co)Promotors are prof.dr. A.D. Nelson and dr. S. Zwart from the Faculty ITC.
Africa’s wetlands are diverse and highly productive ecosystems that hold significant potential to enhance food production. However, sustainable management is essential to mitigate negative impacts from agricultural use. While considerable progress has been made in understanding wetland ecosystems, knowledge gaps persist, particularly for small and intermittent wetlands. Advances in Earth observation (EO), especially through the Copernicus Program, offer opportunities to address these gaps with high-resolution satellite data.
This thesis explores how EO can support sustainable management of African wetlands to improve food security. It investigates aspects of wetland ecosystems and management along the terrestrial-aquatic continuum. Firstly, key requirements were identified and a framework of four information layers designed—Wetland Delineation, Land Use/Land Cover (LULC), Surface Water Occurrence (SWO), and Wetland Use Intensity (WUI)—based on Copernicus imagery. SWO and WUI captured wetlands dynamics, while LULC and WUI revealed drivers like land conversion and intensification. Secondly, the WUI layer was further analysed for Rwanda, demonstrating its utility in monitoring pressures on wetlands at national and local scales. Cloud computing enhanced the efficiency of WUI calculation, increasing applicability for wetland management. Thirdly, Sentinel-2 imagery was used to model water quality by assessing small reservoir turbidity in dammed wetlands in Kenya. The models were calibrated using both laboratory-grade equipment and low-cost sensors as affordable alternatives for in-situ measurements. While the laboratory-based calibrations demonstrated high accuracy, the low-cost sensors achieved moderate agreement. The study underscores the potential of EO-based assessments for efficient turbidity monitoring in small, often unmonitored water bodies. Fourthly, an analysis of small reservoirs in Kenya further explored the use of modelled turbidity time series as an indicator of natural and human impacts on wetlands using machine learning. Key drivers of annual turbidity included windspeed and topography, while long-term turbidity was influenced by land cover and WUI, highlighting the importance of effective land management strategies.
Overall, EO provides relevant information and enables integration across scales, and quantification of interlinkages. Future research should enhance understanding of small, seasonally inundated wetlands and scalable water quality assessments while leveraging high-resolution data, cloud computing, and in-situ measurements to promote sustainable wetland management for improved food security.
