Research Projects 2024-2026

• Natasha Ribeiro (Mozambique, Universidade Eduardo Mondlane)

• Patrick Hostert (Germany, Humboldt-Universität zu Berlin )

Shifting cultivation systems are key for sustaining the food security of rural populations across Africa. Simultaneously, shifting cultivation is driving the clearing of natural vegetation and related environmental trade-offs, such as adverse impacts on biodiversity conservation and carbon storage. Maintaining a balance between rightful land use for semi-subsistence agriculture on the one hand, and biodiversity conservation on the other hand is not trivial because knowledge on the dynamics of shifting cultivation systems remains scarce. This is particularly the case in designated conservation areas and their buffer zones, which are oftentimes biodiversity hotspots where rapid agricultural expansion can severely undermine conservation efforts.

This project aims at developing a transferable workflow to improve our knowledge of shifting cultivation dynamics across three protected areas in Mozambique. Based on time series of Copernicus satellite imagery and state-of-the-art machine learning, we will map yearly indicators of active and fallow cropland extent, and woody vegetation cover. The resulting maps will enable unique insights on the intensity, cyclicity, and trajectories of shifting cultivation and related vegetation clearing and recovery processes over a period of 10 years, which are relevant to maintain a balance between food security on the one hand, and biodiversity conservation on the other hand.

• Abebe Mohammed Ali (Ethiopia,Wollo University )

• María Dolores Raya Sereno (Spain, Spanish National Research Council (CSIC) )

Droughts are devastating natural hazards that directly impact the livelihoods of many regions, especially those located within semi-arid climate regimes. Indeed, the Horn of Africa is one of the most drought-prone and vulnerable regions in the world, with recent drought events severely deteriorating food security in Ethiopia, Kenya and
Somalia. Borena, a largely pastoral region in Southern Ethiopia, was one of the most affected during the recent droughts in 2021 and 2022, degrading livestock health and forcing millions of inhabitants to migrate. The lack of readily available data and up-to-date information impedes local NGOs or authorities to implement mitigation strategies to limit the socio-economic and environmental impacts of these droughts. Satellite or Earth Observation (EO) data has the potential to fill this gap by providing openly-available and timely information in both space and time. Indeed, traditional drought indicators rely solely on meteorological data but these largely fail to account for changes to vegetation health and productivity, which is the most important aspect for farmers and pastoralists. As such, this project aims to integrate state-of-the-art EO datasets, such as hyperspectral and thermal imaging, with physically-based models to better account for water availability and stress using open-source and cloud-based methods.

• Marié Elizabeth Smith (South Africa, Council for Scientific and Industrial Research )

• Dalin Jiang (United Kingdom, University of Stirling )

Water sufficiency is fundamental for human health and economic development especially in countries facing water shortages such as South Africa. This project aims to exploit Sentinel-2 Multispectral Instrument (MSI) high spatial resolution imagery to monitor water quality in small water bodies including dams of the Western Cape. The project will work with key users and groups in S. Africa to co-design the EO-based products.

The most appropriate atmospheric correction model for MSI over these waters will be evaluated and selected based on in situ measurements, towards developing algorithms for accurately estimating water quality including chlorophyll-a (Chl-a) and total suspended solids (TSM) concentrations. An open-source and scalable cloud-based processing chain will be developed to produce time-series water quality datasets from 2015 to 2025 over the region to explore their status and changes, and investigate the potential links between water quality, surface area dynamics and extreme events. Processing chains and dataset will be presented to South African users; these outputs can be used to enhance their capability in EO-based water quality monitoring, to support the sustainable use of water resources in South Africa, and to address the United Nations Sustainable Development Goal 6 and the Africa Union Agenda 2063 Goal 7.

• Zeinab Salah Abdullah (Egypt, Egyptian Meteorological Authority )

• Milica Stojanovic (Spain, University of Vigo )

This research focuses on the northern Mediterranean part of Africa, one of the most water-stressed regions in the world. Drought is a natural phenomenon that particularly impacts this region, with huge consequences on water resources, vegetation, agriculture, population displacement, and human deaths. Thus, the objective of this proposal is to evaluate the occurrence and characteristics of droughts that have affected this region, considering the possible effects of irrigation, temperature, and precipitation, and develop drought hazard maps. The causality of spatial drought and changes in atmospheric and hydrological conditions over the primary production will also be assessed, which will permit the development of a predictive model based on machine learning. This will be possible thanks to ESA/EO data, which provides a consistent and reliable source across a region with a poor weather station network. The combined assessment of drought hazard, vegetation susceptibility, and density of population will allow the development of final maps of drought risk affectation. The outcomes will provide valuable insights into the dynamics of drought and will enhance our understanding of drought causality and its impact on vegetation. Finally, this project will contribute to the development of more effective drought management strategies, supporting food security and environmental resilience.

• Kofi Asare (Ghana, Ghana Space Science and Technology Institute)

• Kristof Van Tricht (Belgium, Vlaamse Instelling voor Technologisch Onderzoek
  (VITO) )

GEOMaize aims to develop end-of-season maize yield estimation in Ghana by integrating existing and new field-level data into a state-of-the-art yield estimation algorithm based on few-shot learning. Maize is chosen due to its significant contribution to Ghana’s agricultural sector, accounting for approximately 23% of the country’s agricultural GDP and serving as a primary food source for many households. Maize plays a crucial role in Ghana’s food security and agricultural economy. Accurate yield estimation is essential for policy-making, resource allocation, and improving food supply chains. Leveraging field data such as a crop type map, crop yield and remote sensing information collected by the Ghana Space Science and Technology Institute (GSSTI) in 2021, 2022, and 2023, and a few-shot learning workflow developed by VITO, this project will build

• Munyaradzi Davis Shekede (Zimbabwe, Zimbabwe National Geospatial and Space Agency )

• Martin de Graaf (The Netherlands, The Royal Netherlands Meteorological Institute (KNMI) )

Zimbabwe is one of the African countries severely affected by high air pollution levels with significant impacts on human health and the environment. To develop mitigation strategies and sustainable policies, monitoring of air quality (AQ) is essential. However, Zimbabwe has few air quality monitoring stations, primarily concentrated in the densely populated industrial area of Harare (Matandirotya et al., 2023). This results in significant data gaps, limited assessment of air quality, ineffective health and environmental policies. Therefore, this proposal seeks to 1) densify air quality monitoring network through installation of AQ sensors; 2) explore the spatial and temporal variations in TROPOMI Aerosol Optical Thickness (AOT) patterns across Zimbabwe and validate AOT products; and 3) develop and evaluate a smoke-dust discriminator for TROPOMI. The project will relate the satellite measurements to the measurements at ground level, so that they can be used to monitor air quality in detail in space and time. The workflow developed within this project will make TROPOMI-
based estimates of aerosol loadings across Zimbabwe available for routine air quality monitoring and as a basis for the calculation of the UN Sustainable Development Goals (SDG) target indicators 11.6.2 and 3.9.1.

• Nadia Ouaadi (Morocco, UMP6P – University Mohammed VI Polytechnic )

• Nicolas Baghdadi (France, INRAE- National Research Institute for Agriculture, Food and the Environment )

Surface soil moisture (SSM) is identified as an essential climate variable crucial for characterizing the
earth’s climate. In the agricultural sector, SSM is playing a vital role in the irrigation scheduling and productivity of crops. This is particularly important for the southern-Mediterranean region, which is suffering from severe water shortages. SSM is typically estimated from microwave remote sensing data, but is provided with either coarse spatial resolution or low temporal frequency. Thermal data, on the other hand, are given with a good trade-off between spatial and temporal resolutions. As they capture emitted thermal radiation, which is influenced by soil water content, thermal data can be used for SSM mapping.
However, the relevance and potential of these data are not yet well explored. The aim of this project is first to downscale Sentinel-3 using Landsat data and fusion methods. Second, the downscaled data will be used for daily mapping of SSM at high spatial resolution. The maps will be compared with SSM maps derived from Sentinel-1 data, and both will be assessed using in-situ measurements. These maps will be made available to researchers, farmers, water managers and policy-makers to enable informed decisions on irrigation strategies, crop selection and land management practices.

• Loudi Yap (Cameroon, National Institute of Cartography (NIC) )

• Makan Karegar (Germany, University of Bonn )

Cameroon’s coastal region is highly vulnerable to flooding and sea-level rise, which affects infrastructure, ecosystems, and local communities. Despite the growing risks from climate change, there is currently no water-level monitoring system along the coast or for the major rivers in the country. In-situ monitoring networks are crucial for calibrating and improving forecast models, validating satellite altimetry data, and developing early-warning flood systems.
At the University of Bonn (Germany), we have developed a cost-effective water-level monitoring sensor using Interferometric Reflectometry (GNSS-IR), known as the Raspberry Pi Reflector (RPR). The RPR provides 1-2 cm accuracy, operates reliably in extreme weather conditions and requires minimal operational costs with just a one-time installation visit. We propose installing 8 RPR units at strategic locations in Cameroon to enhance water level monitoring infrastructure and validate Sentinel-3 and Sentinel-6 satellite altimetry data. The project includes fieldwork and a short visit by Co-PI Dr. Karegar to train personnel at the NIC in Cameroon, with a particular emphasis on a framework following the project’s completion.

• Chiamaka Ehiemere (Nigeria, University of Nigeria, Nsukka )

• Nick Gutkin (Belgium, Vlaamse Instelling voor Technologisch Onderzoek (VITO) )

Rice represents an important staple food in Nigeria, providing nutrition and income to the populace and smallholder farmers. Irregularities and unpredictable weather patterns due to climate change have disrupted crop development, led to moisture shortages, and decreased crop production. To address this challenge, the
federal government developed an early warning drought monitoring system using a single drought index, the Standardized Precipitation Index (SPI). However, SPI is limited as it relies only on precipitation and does not consider other crucial agricultural drought indicators, such as temperature, soil moisture, and plant health. The proposed research aims to expand the existing drought early warning system with indicators derived from multi-source satellite data, including climate, soil moisture, and vegetation greenness. The system will be developed and validated for rainfed rice fields in Nigeria, using local in-situ data. Additionally, estimates of current drought conditions will be combined with climate forecasts using machine learning methods to
forecast agricultural droughts up to 6 months in advance. The targeted drought monitoring and forecasting system is expected to provide relevant and actionable information to all local stakeholders in the agri-food sector, thereby significantly increasing the climate resilience of Nigeria’s rural communities.

• Dr. Ernest William Mauya (Tanzania, Sokoine University of
  Agriculture)

• Sadegh Jamali (Sweden, Lund Unversity)

Estimating forest biomass and carbon stock is crucial for assessing forests’ role as carbon sources or sinks and supporting sustainable forest management to improve food security and sustainability. With increasing concerns over global climate change, there is a growing need for efficient, accurate methods to estimate and report forest biomass and carbon stocks at local, national, and global scales. The West Usambara Mountain forests are one of the key conservation areas in Tanzania that provide essential ecosystem services, including carbon storage and sequestration. Estimating biomass and carbon stocks is crucial for understanding the health of forest ecosystems, which directly supports food security by maintaining soil fertility, regulating water resources, and providing essential services that underpin agricultural productivity. This project aims to enhance biomass and carbon stock estimation in the Usambara forests by integrating multimodal remote sensing data with field measurements. Machine learning models will be used to develop a sustainable monitoring framework for forest resource management and inventory. This framework will enable continuous monitoring of forest health, tracking changes during afforestation, deforestation, and climate change impacts. Additionally, it will help identify vulnerable areas with high reforestation potential, contributing to better forest conservation and climate mitigation efforts, and enhancing food security.