Multi-model Data Warehouse for Multidimensional Analysis of Flavescence Dorée: Application to the Bordeaux Region

Flavescence dorée is a highly contagious grapevine disease that poses a significant economic threat to European viticulture. Its current management relies on symptom monitoring, uprooting infected plants, insecticide treatments, and reservoir management. However, effective disease control is hindered by the disease's asymptomatic period and symptom similarities with Bois Noir, necessitating complex data-driven monitoring. Currently, the reliance on traditional spreadsheet software limits the scalability and spatio-temporal analysis of the massive datasets collected. To address these issues, this paper proposes a Business Intelligence architecture based on Spatial OLAP (OnLine Analytical Processing). By integrating Data Warehousing with Geographic Information Systems, the proposed system enables intuitive, multi-dimensional analysis of diverse data types—including graphs for dispersal modeling and JSON for scientific flexibility. Relying on a Multi-Model Data Warehouse with schema-on-read extensibility, this architecture provides an agile, "drag-and-drop" platform for extension services, farmers and researchers in epidemiology. This approach could facilitate rapid detection, predictive modeling, and optimized resource allocation, ultimately shifting pest control towards a more ecological and economically sustainable model.