NASA Harvest Conducts Advanced Training on Cropland Mapping in Dakar
In a collaborative effort to enhance food security assessments in the region, the University of Maryland and NASA Harvest, led by University of Maryland Assistant Professor Catherine Nakalembe, in partnership with the Centre de Suivi Ecologique (CSE), SERVIR, Direction De L'analyse, De La Prévision Et Des Statistiques Agricoles (DAPSA), and the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), recently concluded an intensive five-day training session in Dakar, Senegal. The training, held from June 3-7 2024, attracted 20 technical analysts from CSE, DAPSA, and ICRISAT who are involved in agricultural monitoring and decision support.
The NASA Harvest team was led by Assistant Professor Catherine Nakalembe, Assistant Professor Hannah Kerner from Arizona State University, School of Computing and Augmented Intelligence, Machine Learning Engineer Ivan Zvonkov, Faculty Specialist Diana Botchway Frimpong, Graduate Research Assistant Adebowale Daniel Adebayor, Undergraduate Research Assistant, Alana Ginsburg, and Kevin Horn, the NASA SERVIR Applied Sciences Team Portfolio Manager providing additional support.
The primary aim of the training was to equip participants with the necessary skills to produce reliable Earth Observation (EO) products for resource management. Specifically, the training focused on teaching analysts cropland mapping and area estimation methodologies using Google Earth Engine (GEE). This capacity development initiative is crucial for supporting food security assessments and informed decision-making in the region.
The training covered a comprehensive curriculum designed to ensure participants could effectively utilize these advanced tools and techniques. This included an in depth session on the use of GEE for generating annual cropland maps, how to critically evaluate new and existing cropland products, producing crop condition assessments using validated crop masks, and how to efficiently compute robust, unbiased annual cropland area estimates.
A highlight of the program was a field excursion on day two, where participants had the opportunity to gain hands-on experience using GoPro helmets for crop type data collection. This practical session was vital for participants to appreciate the effectiveness, the scalability of the process and reliability of the tool.
The General Director of CSE, Dr. Cheikh Mbow, a key invited guest, reiterated the success of the training during his remarks and emphasized the importance of such initiatives in building local capacity and improving agricultural monitoring practices in Senegal. In the plenary and feedback sessions, participants, including Adama Mamadou Sarr, SERVIR Theme Lead - Agriculture & Food Security, expressed their appreciation for the tools and knowledge imparted during the training. They highlighted the practical applicability of the skills learned and the potential impact on their work in food security assessments.
The team also visited the DAPSA headquarters at the new Sphère Ministérielle de Diamniadio, a government complex where the headquarters of the Ministry of Agriculture, Rural Equipment, and Food Sovereignty and other ministries are located. The team had the opportunity to present the objectives of their work to the Director of DAPSA, Dr. Ibrahima Mendy. After learning of the Helmets Crop Labeling Initiative, DAPSA leadership became interested in how the technology could support their own operations. There was also discussion on adding internal GIS capacity within DAPSA.
This training marked a significant step in transferring advanced technological skills to local analysts, fostering greater self-sufficiency in agricultural monitoring and decision support in the region. The collaboration between the University of Maryland, NASA Harvest, and their partners underscores a shared commitment to enhancing food security through innovative and practical solutions. At the end of the training, NASA Harvest Africa Lead Catherine Nakalembe and AI Lead Hannah Kerner, presented five GoPro cameras to the Centre de Suivi Ecologique. These cameras will be instrumental in crop type data collection, enhancing the accuracy and reliability of the crop type maps developed in the future.