Crop planting dates are a dynamic feature of agricultural systems. In the US Corn Belt, planting dates display high spatial and temporal heterogeneity, varying by as much as 2 months. This variation impacts both agricultural productivity as well as the larger human-natural system by influencing the timing of nutrient, carbon, and water cycles. While broad-scale drivers of planting decisions are relatively known, an understanding of fine-scaled drivers and variation has been limited by coarse data availability, generally limited to the state level. This knowledge gap hinders our ability to project how planting dates may shift and adapt to climate change. For example, some studies project crop planting dates will need to shift earlier to best match growing season temperatures under climate change, but co-occurring changes in early season precipitation may inhibit these shifts in poorly drained soils.

Here, we take advantage of Landsat satellite-derived planting date maps to refine our understanding of why and when farmers plant. Focusing on the US Corn Belt between 2000-2020, we investigate historical trends and variation in maize and soybean planting. Our data allows us to link local weather, soil, yields, and crop choices, providing a more full picture of conditions that favor or inhibit planting. Finally, we present a framework under development to formalize these insights in a coupled human-environment systems model to better understand how farmer decisions and adaptive behavior mediates climate change impacts on Great Lakes watersheds.