The project activities started with a thorough analysis of the user requirements and system design. The user requirements were collected both through direct interviews of the users providing their fields for the MAGDA demonstrators and through dedicated surveys. The user requirements analysis included the implications for the MAGDA user interface, based on validated principles for human-computer interaction and interface design guidelines. The system design included various levels: hardware level, data level, data-processing level, and modeling level, as well as the delivery of the results to the end-users by interfacing with a Farm Management System. The system design included the selection of the three demonstrator areas, identified in Italy (Cuneo), France (Burgundy), and Romania (Braila). Site inspections were conducted by the MAGDA partners to select the best solutions for the deployment of the instrumentation, as well as to interact directly with the end-users.
Deployment of nine low-cost, dual-frequency, Galileo-enabled GNSS stations, together with fourteen low-cost in-situ sensor stations was completed. The collection of authorizations needed to fly the meteodrones appropriately to meet the project objectives is being completed.
An evaluation methodology and plan for MAGDA was defined. The plan involved defining the goals of the system, identifying key performance indicators (KPIs), developing a data collection plan, establishing a baseline, evaluating the system's performance, identifying areas for improvement, repeating the process regularly, and communicating the results to relevant stakeholders. Relevant and important KPIs to monitor the system's performance were selected.
The MAGDA data environment was carefully analysed and designed to ensure efficient and proper usage of data and exploitation of results.
Severe weather cases were identified and selected to be used to test the performance of the WRF meteorological model. These case studies have been chosen based on their relevance to improving agricultural forecasting, considering the specific interests of crops in different target sites. The selected case studies were six in total, two for each site. Based on the end-user feedback, the two cases for the Italian site targeted rainfall events, while the four remaining cases targeted hail events.
Water balance simulations were run to form the foundation for an Irrigation Advisory Service. Water balance simulations have been conducted for three pilot areas in France, Italy, and Romania utilizing the SPHY simulation model. The spatial results indicate a generally high correlation between the SPHY model's evapotranspiration output and remote sensing data for the pilot areas in France and Italy, although adjustments to certain input parameters are needed for better performance. The Romanian sites show less correlation.
