Our Expertise

Data Science & Modeling

Improved modeling is a requirement for nearly all of the innovations needed to manage climate risk. Yet, predicting the future site-specific impacts of climate change remains a significant challenge. Traditional catastrophe risk modeling is based on applying past loss data to multiple peril-specific scenarios, but to accurately price policies and send effective risk signals, especially for high-value property and infrastructure, reliable assessments are required for the next 30 to 50 years. With climate change causing changes in the frequency and severity of weather events, actuarial data are of little relevance. Climate impacts are also fundamentally more complex than other perils, as many impacts are interconnected, and extreme weather events affecting one sector at one location can have cascading effects on other sectors and locations, creating challenges for insurers operating in a national or global market.

CIRCAD researchers are connecting recent advances in computing, remote sensing, and climate analytics with parallel advances in explainable machine learning, uncertainty representation, and probabilistic programming. This research will make risk models more defensible, transparent, and adaptable. It will also facilitate inclusion of knowledge and data across disparate domains and geographies.

Examples of research by our faculty on this theme include:

Nature-Based Solutions

Nature-based solutions, aka natural infrastructure, use natural processes and ecosystem services to support engineering objectives, such as reducing flood damages or securing safe and ample water supplies. Examples of nature-based solutions include marshes and dunes that protect coastal property and infrastructure against storms and erosion, as well as forests and wetlands that reduce river flooding and purify water for downstream communities.
Nature-based solutions can play a critical role in enhancing the resilience of human and ecological systems in a changing world (e.g., natural hazards and climate change). Evidence suggests that nature-based approaches can be equally or more cost-effective than conventional structural approaches for flood risk reduction in some situations. Unlike conventional infrastructure, natural infrastructure often has the capacity to recover from extreme weather- and climate-related hazards and self-adjust to ongoing environmental changes.

CIRCAD researchers are developing methods for considering, evaluating, and quantifying the value of nature-based solutions at the scale of individual projects and over large spatial areas at the system-scale.  They are also helping to scale up financing of natural infrastructure projects in the United States by improving assessment of their value in reducing climate risk.

Examples of research by our faculty on this theme include:

Risk Assessment & Decision Theory

Determining appropriate action in the face of the complex, uncertain, and constantly evolving risks of climate change requires fundamental research in decision theory and application. For example, the dynamic nature of the climate change problem necessitates advances in modeling and analyzing decisions under changing conditions. This includes methods for representing the process of learning over time, sequential and adaptive decision-making, and identifying strategies that are robust to shifting constraints and objectives, such as those stemming from changing regulatory requirements. There is also a need for improved approaches for handling deep uncertainty in model-based forecasts as well as methods for making decisions that perform well under uncertainties and variations in model inputs, parameters, and structural assumptions. 

Many households and businesses who would benefit from insurance do not purchase it, while others who have it do not need it. There is also evidence that conventional pricing incentives for implementing risk mitigation measures may not be sufficient to change behavior. Behavioral economics can be used to understand behavior that does not adhere to standard models of choice and decision-making. This can help industry enhance their risk signaling abilities, evaluate the market potential of new products, and incentivize the adoption of resilience-building strategies.

CIRCAD researchers are advancing methods for adaptive learning and updating, decision-making under uncertainty, and use of behavioral decision theory to understand risk perceptions and responses. 

Examples of research by our faculty on this theme include:

Economic & Policy Analysis

The ability of the insurance sector to be a key player in building a climate-resilient society is subject to a variety of policy complexities at multiple levels. 

At the federal level, a growing reliance on government as the insurer of last resort is unsustainable, which will prompt new federal policies to limit private sector market withdrawals. Having the insurance industry be an effective partner in this effort will require productive science-based dialogue around the potential of various policy proposals. Also at the federal level, evolving regulations related to climate risk disclosure and the increased focus on sustainable finance create a situation in which companies that do not adequately account for climate risks may face legal and reputational consequences. Current insurance industry disclosures indicating that their exposure to climate risk can be addressed by simply raising prices are unlikely to be viewed as sufficient, especially given recent market withdrawals. 

At the state level, regulators play a significant role by protecting policyholders, ensuring the fair and efficient functioning of insurance markets, and maintaining the stability of the insurance sector itself. In some states, regulators have successfully preserved stability in the face of growing climate risk while, in others, regulatory distortions in risk-based pricing have contributed to insurers’ decisions to withdraw. It will be important to develop new approaches to state regulation that enable rapid industry innovation in the face of escalating climate risk.

At the local level, the insurance sector, by holding 30% of all municipal bonds, has significant leverage to ensure that municipalities’ investments are consistent with climate resilience. An effective disclosure or rating system will be key to these evaluations. 

Finally, and perhaps unsurprisingly, there is fragmentation of financial regulation across levels and sub-sectors which can obscure the scale of the challenge or delay development of innovative solutions. While there are some mechanisms for greater federal coordination and oversight, these may not always be advantageous or sustainable.

CIRCAD researchers are working to illuminate policy challenges and opportunities, including identification of the critical policy enablers (or inhibitors) to accurate risk-based pricing, market stability, and cooperation regarding risk management interventions and climate data and information sharing. 

Examples of research by our faculty on this theme include:

Insurance & Finance Innovation

Addressing the unique challenges posed by climate change will require innovation in insurance products and risk management strategies. This is because even if precise estimation of future climate risk exposures were possible, and actors were making wise and rational decisions, the economics and politics of insurance markets are unlikely to be conducive to full risk-based pricing of traditional 12-month policies. Relatedly, while individual policyholders may be able to take preventative measures to protect their properties from wind, hail, or interior water damage, they do not exert much personal control over many larger-scale climate risks, such as wildfires, sea level rise, storm surges and heatwaves. These risks need to be managed at the community, county, state, or national level. Thus, there is a need for research to conceptualize, analyze, and model new paradigms for risk management strategies and products that provide coverage to the most vulnerable while being economically and politically sustainable.

CIRCAD researchers are investigating alternative forms of risk management such as layered strategies, community-based catastrophe insurance (CBCI), and parametric insurance.

Community Engagement

Engaging regulators, federal and state policy-makers, local governments, community organizations, and everyday people is essential to meeting all of CIRCAD’s goals. The Nicholas Institute for Energy, Environment, and Sustainability (NIEES) at Duke is a key participant in CIRCAD. NIEES brings connections to a range of governments, institutions, and communities. NIESS has a well-earned reputation for not only delivering timely, credible analyses to these parties, but also for its ability to convene them in developing opportune, effective, and economically practical solutions to critical challenges. For example, Institute staff work with the World Bank, the United Nations, the White House and numerous federal agencies, in addition to state agencies, programs, and communities in North Carolina and across the US.

Similarly, the Institute for Resilient Infrastructure Systems (IRIS) at UGA works in collaboration with government and private sector (for-profit and non-profit) organizations across a broad range of disciplines in social and natural sciences, law and policy, and engineering to support the development of next-generation infrastructure systems – including many that are nature-based – that provide genuine and enduring resilience value.