Local-scale cereal yield forecasting in Italy: Lessons from different statistical models and spatial aggregations

Statistical, data-driven methods are considered good alternatives to process-based models for the sub-national monitoring of cereal crop yields since they can flexibly handle large datasets and can be calibrated simultaneously to different areas. Here, we assess the influence of several characteristics on the ability of these methods to forecast cereal yields at the local scale. We look at two diverse agro-climatic Italian regions and analyse the most relevant types of cereal crops produced. Models of different complexity levels are built for all cases by considering six meteorological and remote sensing indicators as candidate predictive variables. Yield data at three different spatial aggregation scales were retrieved from a comprehensive, farm-level dataset over the period 2001-2015. Overall, our results suggest better predictability of summer crops compared to winter crops, irrespective of the model considered. At higher spatial resolution, more sophisticated modelling techniques resting on feature selection from multiple indicators outperformed more parsimonious linear models. These gains, however, vanished as data were further aggregated spatially, with the predictive ability of all competing models converging at the agricultural district and province levels. Feature-selection models tended to elicit more satellite-based than meteorological indicators and the selected features were, in most cases, equally distributed along the plant growing cycle.

JRC.C.6 - Economics of Climate Change, Energy and Transport