TLDR: This research introduces a new geospatial agent-based model that helps assess climate risks by simulating how economic firms adapt to climate hazards. Using evolutionary learning, firms can adjust strategies like budget allocation and pricing. The model demonstrates that even under severe climate stress (like riverine floods), firms can evolve to maintain production levels similar to a no-hazard scenario over decades, highlighting the importance of adaptive behaviors and revealing systemic risks through supply chain disruptions. The open-source framework provides tools for financial institutions and companies to quantify climate risks and evaluate adaptation strategies.
Climate change poses significant threats to economic systems, with complex interactions between localized hazards and global economic networks. Traditional economic models often struggle to capture these dynamics, particularly the adaptive behaviors of businesses and the cascading effects through supply chains. A new research paper introduces an innovative approach to address these challenges.
The paper, titled “Adaptive Learning in Spatial Agent-Based Models for Climate Risk Assessment: A Geospatial Framework with Evolutionary Economic Agents,” presents a novel geospatial agent-based model (ABM). This framework integrates detailed climate hazard data with evolutionary learning mechanisms, allowing economic agents, specifically firms, to adapt their strategies over time. The lead author of this research is Yara Mohajerani.
The core of this model lies in its ability to simulate how firms evolve their responses to climate stress. It combines a Mesa-based spatial modeling framework with CLIMADA climate impact assessment data. This allows for the creation of a spatial economic network overlaid on climate projection maps, incorporating hazard and vulnerability functions for different locations.
How the Model Works
The model features two main types of economic agents: households and firms. Households supply labor, consume goods, and can relocate based on local hazard conditions. Firms, on the other hand, use labor, capital, and material inputs from suppliers within a global supply chain network. These firms are the key players in the evolutionary learning system.
Firms in the model adapt six crucial strategy parameters through a process of fitness-based selection and mutation. These parameters include: labor budget weight, input budget weight, capital budget weight, risk sensitivity multiplier, price responsiveness factor, and wage adjustment sensitivity. Climate events act as a selection pressure, favoring strategies that enhance survival and growth under stress. This means firms don’t just react to climate events; they learn and evolve their business strategies to become more resilient.
The model tracks each firm’s performance over time, considering factors like money growth, production stability, longevity, and resource balance. Based on a fitness function, successful strategies are propagated, while less successful ones are replaced. This evolutionary process allows for the emergence of diverse, climate-adapted business strategies without explicit programming of every adaptive behavior.
Key Findings and Implications
The researchers demonstrated their framework using riverine flood projections under a high-emission scenario (RCP8.5) until the year 2100, comparing it against a baseline no-hazard scenario. The simulated economy included commodity producers and manufacturers in flood-prone areas, along with households.
Initially, climate hazards caused significant disruption, leading to diminished production, firm wealth, and labor projections. For example, in 2050, average per-firm production was substantially lower under the RCP8.5 scenario compared to the baseline. However, a remarkable finding was that due to evolutionary adaptation, firms were able to converge with or even surpass baseline production levels by the end of the century. This highlights the critical role of adaptive behavior in mitigating the long-term effects of climate shocks.
The study also revealed systemic risks: even firms not directly exposed to floods experienced impacts through supply chain disruptions. Households faced higher unemployment rates and increased inflation, with the average price of goods projected to be 5.6% higher under RCP8.5 by 2100 compared to the baseline. Despite these challenges, the adaptive capacity of firms, such as pre-emptively increasing capital in response to regional risks, helped alleviate some of these issues.
Also Read:
- StefaLand: A New Geoscience Foundation Model for Dynamic Land-Surface Predictions
- Geospatial AI Models Offer New Insights into Urban Heat Islands and Climate Adaptation
Conclusion
This geospatial ABM framework offers a powerful tool for understanding and quantifying climate risks. By incorporating spatial heterogeneity and adaptive behaviors, it provides a more accurate assessment of how economic systems can respond to climate change. The open-source nature of the framework, available at this link, makes it accessible for financial institutions and companies to better assess portfolio climate risks, evaluate cost-effective adaptation strategies, and ultimately build more climate-resilient economic systems.
