Agent-based models to assess the impact of climate change on agricultural landscapes in America's Heartland

Blanca Perez Lapena

Southern Illinois University

It is estimated that the Earth’s climate will considerably change in the near future. At the watershed level, changes in climate and human adaptation are likely to affect crop yields and crop management practices, while the prices that farmers face will be impacted by mitigation policies and global trends in the supply and demand of agricultural products. These local, regional and global drivers will simultaneously impact agricultural landscapes and the quantities and qualities of water. To ensure sustainable management of America’s agricultural Heartland, it is important to better understand these relationships. For this purpose, we analyze possible future agricultural landscapes using Agent-Based modeling techniques. Farmer agents are modeled as risk-averse profit-maximizers, satisficers, and conservationists. Under these parameterizations, agents make decisions on crop rotations, forms of tillage and fertilizer rates based on agricultural prices and returns, crop yields, conservation policies and climate scenarios.  Some agents form expectations by incorporating climate projections in their decision making process while others only consider historic climate information. The simulated agricultural patterns are used as input in the hydrological model SWAT to assess the effects of the changed landscapes on water quantity, water quality, and agricultural production over the watersheds under study. The proposed agent-based model is applied to the Big Creek watershed located in Southern Illinois and characterized by large portions of agricultural land including row crops and rural grassland.