Multi-model and multi-system ensemble assessment to inform adaptation to climate change in agriculture

Abstract

In the Anthropocene, the geological epoch when human activity has started to have a significant impact on the planet's climate and ecosystems, the understanding, forecasting, and treatment of key emerging phenomena is not possible without explicitly including human behavior and responses in models. As additional human and ecological systems are connected, uncertainties across systems cascade and amplify, challenging our forecasting capacities. All the above calls for major renovations of current modeling approaches to better integrate human agency into ensemble experiments, so as to achieve a more accurate characterization of uncertainties and improved assessment of the effectiveness of adaptation and mitigation strategies (UNDRR, 2021). This research proposes a methodology to expand the modular hierarchy of ecological systems adopted in conventional ensemble experiments with socioeconomic systems that represent key aspects of human agency and behavior. The proposed hierarchical framework mimics the structure of conventional ensembles, only in this case a human module is incorporated to account for non-linearities in human agency and their impacts on key socioeconomic variables such as income and employment, and how they affect ecological system dynamics. We illustrate our framework with an application to water resources management in an agricultural river basin in central Spain, the Douro River Basin. The hierarchy uses data of the Global Gridded Crop Models (GGCMs), Global Hydrological Models (GHMs), and Global Circulation Models (GCMs) provided in the framework of the Inter-Sectoral Impact Model Intercomparison Project Protocol 2b (ISIMIP, 2022) to force an ensemble of microeconomic mathematical programming models capable of representing irrigators behavior and adaptive responses. This is done in different stages. In a first step, the ensemble of GHMs and GCMs provides information about water discharge. This information, fed into AQUATOOL, a decision-making system used in the study area, allows us to determine the amount of water available for irrigated agriculture. Furthermore, the ensemble of GGCMs and GCMs inform us about changes in crop production due to climate variations. All these data are then used to drive microeconomic models, which simulate the adaptive responses of irrigators. Through these simulations, we obtain valuable information on profit, employment, and the distribution of crops. The resultant hierarchy of ensembles can be used to explore the consequences of multiple strategies under alternative scenarios and models, while accounting for cascading impacts across ecological and human systems. The result is a large database of simulations in which each simulation represents the socioeconomic and environmental consequences of climate change under one specific scenario and combination of ecological and socioeconomic models—thus thoroughly assessing the fundamental sources of uncertainty and providing comprehensive data to inform the adoption of robust strategies that achieve a satisfactory performance under most plausible futures.