Can Machine Learning Help Human-Water Systems Models “Behave” Better?

Advances in environmental and water systems’ modeling have increasingly allowed to understand and accurately reproduce a variety of physical processes, along with their interactions at various spatial and temporal scales. Anthropogenic activities are major drivers of global change and several studies have conceptualized and investigated the mutual influences between human behaviors and water systems dynamics. Yet, human-water interactions remain oversimplified in many models and processes that support strategic policy design. Here, we argue that advanced data analytics and Machine Learning offer several opportunities to better understand, characterize, and model human behaviors in coupled human-water models. In this talk, we provide an overview of recent advances in data-driven behavioral modelling in human-water systems, targeted at explaining water operators’, water users’, and floodplain residents’ behaviors. With examples and applications from various real-world cases, we will showcase how Machine Learning techniques have been demonstrated to help (i) inferring individual value functions from observed patterns of decisions and modeling dynamic preference evolution in time, (ii) analysing and modelling water use behaviors from high spatiotemporal resolution data, and (iii) exploring behavioral determinants of flood resilience with explainable machine learning. We recognize that this is a rapidly evolving research field, and we will also stimulate discussion around key challenges, including modeling decisions under uncertainty, model explainability, data and computational requirements, the scalability of the models for capturing heterogenous behaviors, and the transferability of behaviors information to contexts with similar behavioral characteristics. Overall, this work contributes to the discussion around which behavioral aspects in human-water systems should be prioritized and which individual/societal behaviors can be better included in research on human-water systems.