A Certification System for Third Party Climate Models to Support Local Planning and Flood Resilience
As the impacts of climate change worsen and become salient to more communities across the country, state and local planners need access to robust and replicable predictive models in order to effectively plan for emergencies like extreme flooding. However, planning agencies often lack the resources to build these models themselves. And models developed by federal agencies are often built on outdated data and are limited in their interoperability. Many planners have therefore begun turning to private-sector providers of models they say offer higher quality and more up-to-date information. But access to these models can be prohibitively expensive, and many remain “black boxes” as these providers rarely open up their methods and underlying data.
The federal government can support more proactive, efficient, and cost-effective resiliency planning by certifying predictive models to validate and publicly indicate their quality. Additionally, Congress and the new Presidential Administration should protect funding at agencies like FEMA, NOAA, and the National Institute of Standards and Technology (NIST) who have faced budget shortfalls in recent years or are currently facing staffing reductions and proposed budget cuts, to support the collection and sharing of high quality and up-to-date information. A certification system and clearinghouse would enable state and local governments to more easily discern the quality and robustness of a growing number of available climate models. Ultimately, such measures could increase cost-efficiencies and empower local communities by supporting more proactive planning and the mitigation of environmental disasters that are becoming more frequent and intense.
Challenge and Opportunity
The United States experienced an unprecedented hurricane season in 2024. Even as hurricanes continued to affect states like Texas, Louisiana, and Florida, the effects of hurricanes and other climate-fueled storms also expanded to new geographies—including inland and northern regions like Asheville, North Carolina and Burlington, Vermont. Our nation’s emergency response systems can no longer keep up—the Federal Emergency Management Agency (FEMA) spent nearly half the agency’s disaster relief fund within the first two weeks of the 2025 fiscal year. More must be done to support proactive planning and resilience measures at state and local levels. Robust climate and flooding models are critical to planners’ abilities to predict the possible impacts of storms, hurricanes, and flooding, and to inform infrastructure updates, funding prioritization, and communication strategies.
Developing useful climate models requires large volumes of data and considerable computational resources, as well as time and data science expertise, making it difficult for already-strapped state and local planning agencies to build their own. Many global climate models have proven to be highly accurate, but planners must often integrate more granular data for these to be useful at local levels. And while federal agencies like FEMA, the National Oceanic and Atmospheric Administration (NOAA), and the Army Corps of Engineers make their flooding and sea level rise models publicly available, these models have limited predictive capacity, and the datasets they are built on are often outdated or contain large gaps. For example, priority datasets, such as FEMA’s Flood Insurance Rate Maps (FIRMs) and floodplain maps, are notoriously out of date or do not integrate accurate information on local drainage systems, preventing meaningful and broad public use. A lack of coordination across government agencies at various levels, low data interoperability, and variations in data formats and standards also further prevent the productive integration of climate and flooding data into planning agencies’ models, even when data are available. Furthermore, recent White House directives to downsize agencies, freeze funding, and in some cases directly remove information from federal websites, have made some public climate datasets, including FEMA’s, inaccessible and put many more at risk.
A growing private-sector market has begun to produce highly granular flooding models, but these are often cost-prohibitive for state and local entities to access. In addition, these models tend to be black boxes; their underlying methods are rarely publicly available, and thus are difficult or impossible to rigorously evaluate or reproduce. A 2023 article in the Arizona State Law Journal found widely varying levels of uncertainty involved in these models’ predictions and their application of different climate scenarios. And a report from the President’s Council of Advisors on Science and Technology also questioned the quality of these private industry models, and called on NOAA and FEMA to develop guidelines for measuring their accuracy.
To address these issues, public resources should be invested in enabling broader access to robust and replicable climate and flooding models, through establishment of a certification system and clearinghouse for models not developed by government agencies. Several realities make implementing this idea urgent. First, research predicts that even with aggressive and coordinated action, the impacts of hurricanes and other storms are likely to worsen (especially for already disadvantaged communities), as will the costs associated with their clean up. A 2024 U.S. Chamber of Commerce report estimates that “every $1 spent on climate resilience and preparedness saves communities $13 in damages, cleanup costs, and economic impact,” potentially adding up to billions of dollars in savings across the country. Second, flooding data and models may need to be updated to accommodate not only new scientific information, but also updates to built infrastructure as states and municipalities continue to invest in infrastructure upgrades. Finally, government agencies at all levels, as well as private sector entities, are already responding to more frequent or intensified flooding events. These agencies, as well as researchers and community organizations, already hold a wealth of data and knowledge that, if effectively integrated into robust and accessible models, could help vulnerable communities plan for and mitigate the worst impacts of flooding.
Plan of Action
Congress should direct the National Institute of Standards and Technology (NIST) to establish a certification system or stamp of approval for predictive climate and weather models, starting with flood models. Additionally, Congress should support the maintenance of these and agencies’ capacities to build and maintain such a system, as well as that of other agencies whose data are regularly integrated into climate models, including FEMA, NOAA, the Environmental Protection Agency (EPA), National Aeronautics and Space Administration (NASA), U.S. Geological Survey (USGS), and Army Corps of Engineers. Congressional representatives can do this through imposing moratoria on Reductions in Force and opposing budget cuts imposed by the Department of Government Efficiency and the budget reconciliation process.
Following the publication of the Office of Science and Technology Policy’s Memorandum on “Ensuring Free, Immediate, and Equitable Access to Federally Funded Research,” agencies that fund or conduct research are now required to update their open access policies by the end of 2025 to make all federally funded publications and data publicly accessible. While this may help open up agency models, it cannot compel private organizations to make their models open or less expensive to access. However, federal agencies can develop guidance, standards, and a certification system to make it easier for state and local agencies and organizations to navigate what’s been called the “Wild West of climate modeling.”
A robust certification system would require both an understanding of the technical capabilities of climate models, as well as the modeling and data needs of resilience planners and floodplain managers. Within NIST, the Special Programs Office or Information Technology Laboratory could work with non-governmental organizations that already convene these stakeholders to gather input on what a certification system should consider and communicate. For example, the Association of State Floodplain Managers, American Flood Coalition, American Society of Adaptation Professionals, and American Geophysical Union are all well-positioned to reach researchers and planners across a range of geographies and capacities. Additionally, NIST could publish requests for information to source input more widely. Alternatively, NOAA’s National Weather Service or Office of Oceanic and Atmospheric Research could perform similar functions. However, this would require concerted effort on Congress’s part to protect and finance the agency and its relevant offices. In the face of impending budget cuts, it would benefit NIST to consult with relevant NOAA offices and programs on the design, scope, and rollout of such a system.
Gathered input could be translated into a set of minimum requirements and nice-to-have features of models, indicating, for example, proven accuracy or robustness, levels of transparency in the underlying data or source code, how up-to-date underlying data are, ease of use, or interoperability. The implementing agency could also look to other certification models such as the Leadership in Energy and Environmental Design (LEED) rating system, which communicates a range of performance indicators for building design. Alternatively, because some of the aforementioned features would be challenging to assess in the short term, a stamp of approval system would communicate that a model has met some minimum standard.
Importantly, the design and maintenance of this system would be best led by a federal agency like NIST, rather than third-party actors, because NIST would be better positioned to coordinate efficiently with other agencies that collect and supply climate and other relevant data such as FEMA, USGS, EPA, and the Army Corps of Engineers. Moreover, there are likely to be cost efficiencies associated with integrating such a system into an existing agency program rather than establishing a new third-party organization whose long-term sustainability is not guaranteed. The fact that this system’s purpose would be to mediate trustworthy information and support the prevention of damage and harm to communities represented by the federal government also necessitates a higher level of accountability and oversight than a third-party organization could offer.
NIST could additionally build and host a clearinghouse or database of replicable models and results, as well as relevant contact information to make it easy for users to find reliable models and communicate with their developers. Ideally information would be presented for technical experts and professionals, as well as non-specialists. Several federal agencies currently host clearinghouses for models, evidence, and interventions, including the Environmental Protection Agency, Department of Labor, and Department of Health and Human Services, among many others. NIST could look to these to inform the goals, design, and structure of a climate model clearinghouse.
Conclusion
Establishing an objective and widely recognized certification standard for climate and weather models would support actors both within and outside of government to use a growing wealth of flooding and climate data for a variety of purposes. For example, state and local agencies could more accurately predict and plan for extreme flooding events more quickly and efficiently, and prioritize infrastructure projects and spending. And if successful, this idea could be adapted for other climate-related emergencies such as wildfire and extreme drought. Ultimately, public resources and data would be put to use to foster safer and more resilient communities across the country, and potentially save billions of dollars in damages, clean up efforts, and other economic impacts.
This action-ready policy memo is part of Day One 2025 — our effort to bring forward bold policy ideas, grounded in science and evidence, that can tackle the country’s biggest challenges and bring us closer to the prosperous, equitable and safe future that we all hope for whoever takes office in 2025 and beyond.
PLEASE NOTE (February 2025): Since publication several government websites have been taken offline. We apologize for any broken links to once accessible public data.
The federal government can support more proactive, efficient, and cost-effective resiliency planning by certifying predictive models to validate and publicly indicate their quality.
The cost of inaction is not merely economic; it is measured in preventable illness, deaths and diminished livelihoods.
Through investments in infrastructure for heat safety, Congress can save lives, protect the economy, and enhance resilience nationwide.
A deeper understanding of methane could help scientists better address these impacts – including potentially through methane removal.