Introduction Agricultural yields are increasingly threatened by extreme weather events, yet traditional meteorological indicators fail to capture whether events that cause significant crop losses. We want to address this by developing impact-oriented weather indicators calibrated directly to yield outcomes. Building on the ModOKlim modelling framework, the project aims to produce regionally adapted, policy-ready indicators. Methodologies Yield data across multiple crops is aggregated into a Crop Portfolio Yield (CPY) measure. Nine seasonal indicators covering drought, waterlogging, and heat are then linked to CPY losses via a linear probability model with regional fixed effects (LPMFE). F-score optimisation derives a region- specific threshold, producing a Weighted Index Score (WIS) to trigger early warnings or ex-post assessments. Output - Early warning system (ex-ante): seasonal risk probability maps during the vegetation phase - High-risk region identifier (ex-post): index-based county flags for disaster aid or insurance - Optimised crop portfolio recommendations under future climate scenarios Impact MonOKlim will deliver regionally adapted, policy-ready indicators that support evidence-based decision-making for agricultural policy, including monitoring of the German Climate Adaptation Strategy (DAS) and the climate damage registry (Klimaschadenkataster). The ex-post indicator can simplify disaster relief processing, as counties flagged by the index may qualify for aid without requiring farm-level proof of loss. What I can offer We can share impact-oriented weather indices and the underlying methodological framework (CPY, LPMFE, F-score), county-level crop yield data for Germany across major arable crops, and seasonal risk maps and early warning outputs applicable to insurance, policy, and farm advisory contexts. What I need We are looking for farm-level data such as soil type, or irrigation information to improve model precision, as well as new weather data or novel indicators beyond the current set. Data from other EU countries would enable geographic extension of the framework, and we welcome contact with comparable early warning approaches to benchmark and validate against.

Source: Annual AgData Use Case Summit 2026 booklet.