Skip to main content
main content, press tab to continue
Article | WTW Research Network Newsletter

A guide for navigating the complex landscape of weather and climate research

By Daniel Bannister , Katherine Latham , Álvaro Linares Fuster and Ester Calavia Garsaball | February 5, 2025

If you’re less familiar with climate research, here are a few tips to identify credible sources and make informed business decisions confidently.
Climate|Corporate Risk Tools and Technology|Risk and Analytics
Climate Risk and Resilience

In tackling today’s climate risks, credible research and data are no longer luxuries – they are essential tools for making informed business decisions. Yet access to data alone isn’t enough; it must be interpreted and applied effectively to be actionable. At WTW, with a track record of 20+ years of working with climate scientists, we believe that a transdisciplinary approach – bringing together experts across fields to connect research with real-world applications – is key to making climate data not only accurate but also accessible and actionable.

Recent reports from the Royal Meteorological Society also highlight that collaboration between the insurance industry and academia is essential for translating complex data into practical solutions for adapting to climate change.[1] By collaborating with respected institutions, industry partners can rigorously evaluate models and data, develop techniques like storyline development and scenario analyses, and gain a more comprehensive view of climate risk. For example, expert panels comprising leading specialists in their field offer valuable insights where data may be limited, refining risk assessments and improving the accuracy of predictions. Similarly, placements and internships from academia bring fresh perspectives, helping industry stay in step with the latest trends in climate science.

While academia-industry collaboration improves our understanding of what makes research credible – and helps us spot potential pitfalls – knowing how to evaluate research independently is just as critical. In the sections below, we outline the key qualities to look for to spot credible research, cautionary signs to consider, and the latest techniques in this rapidly evolving field.

Understanding good research and data

Good research and data have four key traits:

  1. Credibility and expertise: Trustworthy research and data often come from well-known authorities, institutions or collaborations in climate science and environmental research. Evaluate the source of data. Are they reputable organizations like NOAA or the IPCC? These institutions have a history of providing reliable and scientifically robust information. Additionally, engage with institutions and networks that specialize in climate science and risk. Organizations, such as the Royal Meteorological Society, not only publish innovative research but also host seminars and events for both experts and non-experts, offering valuable opportunities for learning and collaboration.
  2. Accessibility and transparency: Climate data is now more accessible than ever before. But access alone isn’t enough. High quality research clearly outlines its methods, data sources, and assumptions. Do they use the latest climate model data and methods that are widely accepted in the scientific community? If the information comes from a website, media, or social media, it should link back to the original research. This means that the information can be traced back to its source and verified for credibility.
  3. Peer review and publication: Research that is peer-reviewed and published in reputable journals is usually more credible. Studies in journals like the Journal of Climate or Nature Climate Change are known for their rigorous peer-review processes. Similarly, verify if data from other sources has been reviewed or validated by credible experts.
  4. Relevance, timeliness and application: While data and research may meet the above three points, it’s also essential to consider the recency and context of its application. Climate science advances rapidly, and even highly reputable sources, like IPCC reports, become quickly outdated. Additionally, credible data may be misinterpreted if used outside its intended scope. This is where collaboration with academia is particularly valuable; it can help clarify exactly what the data represents, its limitations, and the appropriate contexts for its use. Look for research that includes clear, actionable recommendations and is explicitly suited to the decisions you need to make. This can help develop targeted risk strategies and tools that are both reliable and practical for assessing climate risks.

Research to be cautious about

But not all research is helpful. Be wary of:

  1. Lack of transparency: Research that doesn’t disclose its methods, data sources, or conflicts of interest should be approached with caution. Studies that only provide summaries without detailed methods or datasets are less reliable and harder to validate. Research that also makes broad claims without supporting data or that cannot be replicated should be treated cautiously. Reports claiming significant climate impacts without regional or sector-specific details are often less reliable and useful.
  2. Questionable accuracy and coverage: With the recent boom of climate analytics solutions, it is common to see products and data claiming overwhelming accuracy and economy-wide coverage. Always question data sources and the associated uncertainties to avoid spurious accuracy in your decision-making and consider the learnings of stablished industries. For instance, the (re)insurance industry has been pricing and managing physical risks for several decades and has validated frameworks to quantify the effects of climate change.
  3. Uncertain funding sources: Consider who is funding the research and their motivations. Research funded by entities with a financial stake in the outcomes might lack objectivity.
  4. Limited peer review: Findings that haven’t been peer-reviewed or are published in lesser-known journals, newspapers or magazines may lack rigor and credibility. Such sources often include opinion or editorial pieces rather than research. Whenever possible, trace information back to its primary source to ensure it has undergone a robust peer-review process and meets high standards of reliability.

Emerging research and techniques

Understanding what makes research good is crucial, but staying informed about new trends and techniques in climate risk assessment is equally important. While these advancements offer significant potential, they also come with challenges to be aware of. Key developments to watch include:

  1. Artificial Intelligence (AI) and Machine Learning (ML): Applying AI and ML to analyze vast amounts of climate and weather data could result in improved risk prediction and deeper insights into climate patterns. However, these models can act as “black boxes” if their internal logic is opaque, raising concerns about reliability and interpretability, particularly in the context of critical climate decisions. Academia-industry collaborations like the UKRI Centre for Doctoral Training in the Application of Artificial Intelligence to the study of Environmental Risks (AI4ER), led by the University of Cambridge, are addressing this challenge by developing transdisciplinary expertise focused on specific environmental challenges. Their aim is to advance the transparency and applicability of AI for environmental risk, ultimately enhancing our ability to understand, monitor and predict environmental risks.
  2. Climate storylines and scenario planning: Visualizing future states through narrative contexts, such as storylines and scenario planning, may aid in strategic decision-making by making complex climate data more accessible and understandable. To read more about how storylines can help assess compounding climate risks and explore contrasting disaster outcomes, see our recent articles on natural catastrophe risks in Australia and wildfire management in Hawaiʻi.[2] [3]
  3. Communicating uncertainty: Climate data and research often involve navigating a maze of choices – deciding which model, scenario, and time horizon to use, alongside the uncertainty in the parameters themselves. New methods, such as ensemble modeling and probabilistic approaches, can help provide clearer confidence bounds, making this inherent uncertainty more manageable and actionable. Effective communication is key, ensuring that decision-makers understand the nature of these uncertainties and can incorporate them meaningfully into risk management. Academic placements and studentships play a crucial role here, equipping the next generation of experts with the skills to communicate complex scientific findings and engage with end-users, including the (re)insurance sector, to ensure research has real-world impact.
  4. Expert panels: Expert panels, composed of specialists, provide valuable insights on complex issues, particularly when data is limited. These panels may play a crucial role in informing risk assessments and strategies, especially in areas with high uncertainty or emerging risks. By pooling expertise, they help clarify complex questions – such as those surrounding uncertainty (as highlighted above) – and guide the interpretation and practical application of data. This collaborative approach ensures that data reflects both scientific and industry perspectives, enhancing its relevance and utility in decision-making.
  5. Exposure and vulnerability data: A complete view of climate risk requires not only hazard data but also detailed information on exposure and vulnerability, which is often outdated or oversimplified. While hazards may be well-modeled, factors such as infrastructure age, building standards, and flood defences also shape a location’s risk. Emerging technologies like remote sensing and AI offer ways to improve the accuracy and quality of exposure data, even in regions with limited observations. Vulnerability is especially complex; buildings may respond differently to the same hazard due to varying construction types, and damage from one event can increase vulnerability to subsequent ones. In regions like the US and Europe, extensive historical claims data underpins well-defined vulnerability curves, but such data is often scarce in other areas. Thorough literature review, drawing on established methodologies and regional studies, can help fill these gaps and forms the foundation for developing the vulnerability functions. By advancing exposure and vulnerability research, we can achieve more accurate and actionable climate risk models that reflect both the physical hazards and societal factors at play.
  6. High-resolution climate modeling: Each new generation of climate models offers finer details, which can improve predictions for extreme events. However, higher resolution requires more computing power and can introduce new uncertainties, especially in areas with little observational data for model validation. It’s important to recognize that more detail does not always mean more accuracy; high-resolution models must be applied with expertise and balanced with other approaches to avoid overconfidence in results.
  7. Integration of climate and catastrophe models: Traditional catastrophe models have been built on historical event data, which can limit their effectiveness in projecting future climate risks. New research and the growing use of climate projections in catastrophe models will allow us to capture a broader range of potential impacts and account for future climate change. We also need forward-looking datasets on exposure (e.g., land use changes, population growth) and vulnerability (e.g., evolving building standards, aging infrastructure) to reflect potential future states. Relying solely on current datasets can result in incomplete risk assessments. Together, these advancements will enhance our ability to predict and manage climate-related risks, creating more resilient planning frameworks for businesses and insurers.
  8. Seasonal predictions: Seasonal predictions have long been used in sectors like agriculture and energy but have yet to gain traction in insurance and risk management due to challenges in translating predictions into reliable short-term planning metrics or long-term investment strategies. A key factor is the alignment of forecast relevance and timing; while predictions are often most accurate at specific times of the year, they may not coincide with critical business cycles, such as January insurance renewals, leading to a mismatch in utility. But sub-seasonal to decadal predictions are an active area of research, and advancements in AI are expected to expand the field further. For instance, partnerships like WTW’s collaboration with the University of Colorado are advancing seasonal climate prediction methods to address these challenges.[4] By using forecasts of El Niño and La Niña events – natural Pacific Ocean fluctuations that significantly impact global weather patterns – companies can anticipate climate-driven risks like droughts, wildfires, and hurricanes months in advance.
  9. Transition and liability risks: Compared to physical risks, the assessment of transition and liability risks associated with climate change is still emerging, largely due to the complexity of modeling economic shifts and regulatory changes. Transition risks – arising from the shift to a net-zero economy – require advanced methodologies that can account for policy developments, technological advances, and shifting market demands. However, the modeling of these risks is still in its early stages, and more research is essential to refine approaches that effectively integrate lessons from climate science into economic projections. This is an area to watch, as advancements in modeling techniques, supported by academic and industry collaboration, will be crucial in equipping businesses to understand and prepare for the financial impacts of the transition to a low-carbon economy. 

References

  1. Royal Meteorological Society. Insurance Industry Stability At Risk Without Stronger Collaboration On Climate Change, New Reports Warn. (2024). Return to article
  2. WTW. A storyline approach for navigating complex and compound climate risks in Australia. (2024). Return to article
  3. WTW. Narrative analysis of wildfire disasters: A case study of Hawai‘i’s 2018 and 2023 fires. (2024). Return to article
  4. WTW. WTW launches partnership with the University of Colorado Boulder to harness the climate prediction revolution.  (2024). Return to article

Authors


Weather & Climate Risks Research Lead
WTW Research Network
email Email

Natural Catastrophe Analytics Manager
email Email

Director – Climate Analytics
email Email

Director, Natural Catastrophe & Climate Risk
email Email

Contact us