Benchmarking the State of AI for Crop Yield Forecasting: A Global Assessment Across Modeling Paradigms
Organizations: Artificial Intelligence Group, Wageningen University and Research, Wageningen, the Netherlands · Earth Observation for Agroecosystems, Agroscope, Zurich, Switzerland · Potsdam Institute for Climate Impact Research (PIK), Germany · Department of Geography, Ludwig-Maximilians-Universität Munich, Munich, Germany ·
Abstract
Accurate crop yield forecasting is essential for food security, agricultural policy, and commodity markets. Although statistical, process-based, and machine learning approaches have all been proposed, their relative strengths remain poorly understood due to the lack of large-scale, standardized comparisons. We address this gap through a comprehensive benchmark of five modeling paradigms -- statistical baselines, process-based models, conventional feature-engineered machine learning, deep sequence models, and tabular foundation models -- using the CY-Bench benchmark spanning 63 country-crop datasets (>12,000 administrative regions) across six continents. Beyond conventional pooled accuracy metrics, we introduce a multi-dimensional evaluation framework that separately quantifies overall, spatial, temporal, and anomaly prediction skill. Across multiple crops and countries, feature-engineered machine learning and tabular foundation models achieve the strongest overall predictive performance. Data-driven models reduce NRMSE in 72% of countries growing maize and 78% of countries growing wheat, with median reductions of 8.8% and 11.9%, respectively. Across all modeling paradigms, spatial prediction is substantially easier than temporal and anomaly prediction. While data-driven models markedly improve temporal and anomaly prediction over statistical baselines, accurately forecasting year-to-year yield variability remains challenging. These findings suggest that future progress in large-scale crop yield forecasting will depend on the improved characterization of the environmental and management factors driving inter-annual yield variability, founded on effective representations for crop yield forecasting systems that support agricultural decision-making, early-warning applications, and food security planning
Figures & tables
| Overall | Spatial | Temporal | Anomaly | |||
| Family | Model | NRMSE | ||||
| Maize (36 countries; 8,211 regions) | ||||||
| Baseline | Trend | 25.4 [19.7, 33.7] | 0.18 [-0.33, 0.54] | 0.71 [0.58, 0.82] | -0.27 [-0.53, 0.07] | -0.10 [-0.17, 0.02] |
| Process-based | LPJmL | 34.1 † [23.5, 50.5] | -0.31 † [-1.05, -0.05] | 0.49 † [0.37, 0.65] | 0.22 ∗ [0.04, 0.49] | 0.08 ∗ [-0.01, 0.28] |
| Feature ML | RF | 24.8 [17.9, 35.3] | 0.25 [-0.24, 0.52] | 0.72 [0.46, 0.82] | 0.27 ∗ [-0.09, 0.50] | 0.16 ∗ [-0.02, 0.25] |
| Foundation | TabICL | 24.2 ∗ [17.0, 33.8] | 0.25 ∗ [-0.13, 0.52] | 0.73 [0.56, 0.83] | 0.33 ∗ [0.07, 0.61] | 0.14 ∗ [0.04, 0.33] |
Appendix figures & tables1 asset
Supplementary material from the paper’s appendix.
Appendix
| Overall | Spatial | Temporal | Anomaly | |||
| Family | Model | NRMSE | ||||
| Maize (36 countries; 8,211 regions) | ||||||
| Baseline | Trend | 0.0 [0.0, 0.0] | 0.00 [0.00, 0.00] | 0.00 [0.00, 0.00] | 0.00 [0.00, 0.00] | 0.00 [0.00, 0.00] |
| Process-based | LPJmL | N/A | N/A | N/A | N/A | N/A |
| Feature ML | RF | 0.3 [-0.7, 1.3] | 0.01 [-0.03, 0.09] | 0.00 [-0.04, 0.02] | 0.06 [-0.10, 0.16] | 0.01 [-0.05, 0.07] |
| Foundation | TabICL | 0.4 ∗ [-1.4, 2.4] | 0.03 ∗ [-0.07, 0.10] | 0.03 [0.00, 0.07] | 0.12 ∗ [-0.04, 0.26] | 0.03 [-0.03, 0.15] |
Crop Yield Prediction
Crop yield prediction estimates harvests from crop, weather, soil, satellite, and management data to support agricultural planning.
See all 21 papers