EUPHORIA-AI: A Three-Layer Satellite-HAP-LAP Observation Architecture for Agricultural Forecasting, Climate-Driven Suitability Mapping, and Long-Term Yield Prediction Across Europe | RISE Research

This paper proposes EUPHORIA-AI, a three-layer observation architecture combining satellite, high-altitude platform (HAP) and low-altitude platform (LAP) data for agricultural forecasting across Europe. The central idea is that no single observation tier delivers both the spatial coverage and the fine resolution that precision agriculture needs, so the system fuses remote-sensing inputs at different altitudes to produce richer, more reliable agricultural intelligence. The architecture targets three linked tasks: short-term agricultural forecasting, climate-driven suitability mapping that identifies where particular crops can be grown as conditions shift, and long-term yield prediction. To do this it draws on variables such as land surface temperature, evapotranspiration, soil moisture, leaf area index and gross primary productivity, integrating satellite-UAV-drone data fusion with deep-learning prediction models and drought and yield models. By layering observation sources and applying machine learning across them, the framework aims to improve the accuracy and lead time of forecasts that matter for food security and farm management at a continental scale. The paper sits at the intersection of aerospace engineering, computer science, sustainable agriculture and data science, and its focus is on demonstrating how a coordinated multi-tier sensing system could move beyond the limitations of conventional single-source remote sensing. Because the system is presented as a proposed architecture, the emphasis is on the design logic, the data pipeline and the suitability of each modelling component rather than on a fully deployed and validated operational result, while pointing toward climate-resilient agricultural planning as the long-term payoff. By treating multi-altitude sensing and machine learning as a single integrated pipeline, the paper makes the case that coordinated observation systems, rather than isolated satellites or drones, are the more promising route to dependable, continent-scale agricultural forecasting under a changing climate.