Putting humans in the loop: coupling behavioral modeling with natural systems' models

Natural systems’ models have done tremendous progress in accurately reproducing a large variety of physical processes both in space and time. Conversely, despite human footprint is increasingly recognized as a major driver of undergoing global change, human behaviors and their interactions with natural processes still remain oversimplified in many models supporting strategic policy design. Recent years have seen an increasing interest and effort by scientists in quantitatively characterizing the co-evolution of nature and society. Nevertheless, state-of-the-art models often relies on behavioral rules empirically defined or derived by general social science or economic studies, which lack proper formalization for the specific case study as well as validation against observational data.

In this talk I will discuss my experiences in modeling human behaviors by taking advantage of the unprecedented amount of information and data nowadays available and of the improvements in machine learning and optimization algorithms. The resulting decision-analytic behavioral models flexibly blend descriptive models, which derive if-then behavioral rules specifying human actions in response to external stimuli, and normative models, which assume fully rational behaviors and provide optimal decisions maximizing a given utility function, where the ultimate goal is not to support optimal decisions but, rather, to understand and model human decisions and behaviors at different spatial and temporal scales.

A number of real world examples in the water domain will be used to provide a synthesis of recent advances in behavioral modeling and to stimulate discussion on key challenges, such as the role of individual behavioral factors in modeling decisions under uncertainty, the scalability of the models for capturing heterogenous behaviors, the definition of model’s boundaries, the identification of behavioral preferences in terms of tradeoff among multiple competing objectives and the dynamic evolution of this tradeoff driven by extreme hydroclimatic events.