Climate Data Gaps Assessment

This document assesses the limitations of climate data available as of late 2024 for electricity sector planning and operations. Climate READi acknowledges that research and data development is rapidly evolving in this space. We hope that this resource, informed by EPRI’s expertise in power system applications, can help focus these activities on high value data development and may consider updating it periodically as work advances.

This report is structured as an outline to systematically assess data gaps by product type (e.g., weather and climate observations, climate model simulations) and details data gaps specific to key climate variables and hazards. In addition, this assessment discusses next steps and considerations for addressing climate data gaps. Key takeaways from this assessment include:

• Observational data gaps of climate variables such as solar irradiance and hub-height wind speeds exist due both to the small volume of historical measurements and to the lack of public access to the proprietary measurements that do exist.
• Data gaps of climate model simulations of variables at spatial (e.g., local) scales are mostly driven by technical and re­source constraints, such as the ability to accurately represent complex local topography, computing speed, and storage space. Similar constraints are often cited regarding provision of relevant temporal (e.g., hourly) data, but are no longer limiting factors in this case. Rather, misalignment between the climate modeling and power sector communities around data use cases and reconciling what data are needed with what data are possible are predominantly driving temporal gaps.
• Observational data on severe weather such as thunderstorms, tornadoes, hail, ice storms, and local scale extreme wind events that are highly impactful to the power system, are limited by the sparse sensor network and its short history. Climate model data for these phenomena are limited by their relatively coarse spatial and temporal resolution. Many extreme weather events can only be resolved by high-resolution weather forecast models. Further, even with unlimited computational resources, these models are still limited in their ability to accurately represent important environmental process, especially at small scales, and become increasingly difficult to validate at fine scales due to a lack of observational data for validation. Lastly, weather forecast models cannot yet be run out for many decades into the future, making for­ward-looking information on individual events difficult to produce. However, climate projections of conditions associated with severe weather can provide some understanding of changing risks.
• Efforts to improve climate literacy in the power sector are valuable to allow those utilizing climate data in downstream power system modeling and analysis tasks to better understand available climate data and potential data applications.
• Collaborations between climate data providers and power system modelers and planners are ongoing and needed to en­sure key data needs are addressed and data are provided in a usable format.
• Notwithstanding the former two bullets, it must be understood that climate data comes from a diverse range of observa­tional and modeling platforms many of which are highly complex. Understanding characteristics like bias and uncertainty of data from different sources across different types of weather and/or electric system conditions is crucial. It is important to consider these factors within the context of specific use cases. For example, resource adequacy assessment requires a good understanding of the data in the tails of weather variable distributions. Organizations seeking to assess climate risks should proactively engage relevant climate specialists rather than simply taking data/tools off the shelf and relying on a non-specialist to analyze them.