Abstract
Spatial information on yield potential is key to determine crop production potential on existing cropland. Although statistical methods are widely used to estimate yield potential and yield gaps at regional to global levels, a rigorous evaluation of their performance is lacking. Here, we compared outcomes from three common statistical approaches against those derived from a ‘bottom-up’ approach based on crop modeling and local weather and soil data for major crops in the United States. Our analysis revealed that statistical methods failed to capture the spatial variation in yield potential, consistently under- or over-estimating yield gaps across various regions. The statistical methods led to conflicting results for decision-making, with production potential almost doubling from one method to another. Given these limitations, we advocate for the use of well-validated process-based crop models coupled with local data and robust spatial frameworks, which provide a more reliable assessment of crop production potential from local to regional scales.