Case study: Climate risk analytics & scenario modeling — a city or utility pilot and the results so far

In 2022, Miami-Dade County launched one of the most ambitious municipal climate risk analytics programs in the United States. Facing $35 billion in assessed property value within FEMA-designated flood zones and a coastline where king tide events had increased 400% since 2006, the county partnered with First Street Foundation and Jupiter Intelligence to build a parcel-level climate risk assessment covering all 900,000+ properties in the jurisdiction. The project moved from procurement to operational deployment in 18 months, produced risk scores now embedded in the county's capital planning and zoning decisions, and generated measurable outcomes that other municipalities and utilities have begun replicating. This case study examines the design choices, implementation challenges, measured results, and transferable lessons from what has become a reference implementation for municipal climate risk analytics.

Why It Matters

Local governments and utilities sit at the operational front line of climate impacts. They own and maintain the physical infrastructure (roads, drainage systems, water treatment plants, electrical substations) that climate hazards damage first. They make land-use decisions that determine whether development occurs in high-risk areas. They set building codes that determine structural resilience. And they bear the fiscal consequences when disaster strikes: FEMA data shows that the average US county now spends 14% more on disaster-related costs than it did a decade ago, adjusted for inflation.

Yet most municipalities make these decisions using backward-looking data. Flood maps based on historical rainfall patterns. Capital plans built around 50-year return period assumptions that no longer hold. Zoning codes designed for a climate that is rapidly changing. The gap between available climate intelligence and decision-making needs has widened as climate impacts accelerate. Between 2020 and 2025, the US experienced 98 weather and climate disasters exceeding $1 billion in damages, compared to 67 in the preceding five-year period.

Climate risk analytics and scenario modeling tools attempt to close this gap by translating global climate model outputs into localized, asset-level risk projections. The technology has matured rapidly: the market for climate risk analytics platforms grew from $1.2 billion in 2022 to $3.8 billion in 2025, driven by regulatory requirements (including TCFD-aligned disclosure, EU Taxonomy eligibility, and SEC climate rules) and growing insurance market disruptions. But deployment at the municipal level remains uneven. A 2025 survey by the National League of Cities found that only 23% of US cities with populations over 100,000 had implemented quantitative climate risk assessments, despite 78% identifying climate adaptation as a top-three budget priority.

The Pilot: Design and Implementation

Problem Definition

Miami-Dade County identified three interconnected challenges that existing tools could not address. First, the county's flood risk maps, last updated by FEMA in 2017, did not account for compound flooding (simultaneous coastal storm surge and inland rainfall flooding), which had caused the most damaging events in recent years. Second, capital planning for drainage infrastructure relied on intensity-duration-frequency curves derived from historical rainfall data that underestimated extreme precipitation events by 15-30% based on observed trends. Third, the county lacked a systematic way to evaluate how different development and infrastructure investment scenarios would perform under future climate conditions.

Technology Selection

The county evaluated five climate risk analytics platforms through a competitive procurement process, ultimately selecting a dual-vendor approach. First Street Foundation provided parcel-level flood, heat, and wind risk scores calibrated to peer-reviewed climate projections (CMIP6 ensemble outputs downscaled to 30-meter resolution). Jupiter Intelligence provided forward-looking hazard analytics for infrastructure planning, including probabilistic projections of compound flooding, extreme heat, and hurricane wind speeds under SSP2-4.5 and SSP5-8.5 scenarios at 5-year intervals through 2070.

The selection criteria weighted three factors heavily: spatial resolution (parcel-level or better), temporal granularity (projections at decade intervals rather than single end-of-century snapshots), and integration capability with the county's existing GIS and asset management systems (Esri ArcGIS Enterprise and IBM Maximo). The winning vendors demonstrated the ability to deliver risk data through API endpoints compatible with the county's enterprise architecture rather than standalone dashboards.

Data Integration

The most time-consuming phase was data preparation and integration. The county needed to align climate hazard layers with its parcel database (900,000+ records), infrastructure asset registry (12,400 assets including roads, bridges, drainage structures, and buildings), and financial systems (capital improvement program and insurance records). Key data challenges included:

Inconsistent geospatial referencing across county departments, requiring six months of spatial data reconciliation. Approximately 8% of infrastructure assets lacked precise GPS coordinates and had to be manually geocoded.

Missing elevation data for 22% of parcels in western unincorporated areas, which required supplementary LiDAR acquisition at $45 per square mile.

Legacy asset management records with incomplete condition assessments, making it difficult to evaluate vulnerability (as distinct from hazard exposure). The county invested $380,000 in field assessments to fill critical gaps for the 2,100 highest-priority infrastructure assets.

Implementation Timeline

The project followed a phased approach. Phase 1 (months 1-6) covered procurement, data acquisition, and platform configuration. Phase 2 (months 7-12) focused on model validation, calibration against observed flood events (Hurricane Irma 2017 and Tropical Storm Eta 2020), and initial risk scoring for all parcels and infrastructure assets. Phase 3 (months 13-18) delivered integration with county planning and capital budgeting workflows, staff training, and public-facing risk communication tools.

Total project cost for the first three phases was $4.2 million, funded through a combination of a HUD Community Development Block Grant for Disaster Recovery ($2.1 million), the county's Stormwater Utility Fund ($1.4 million), and an EPA Climate Pollution Reduction Grant ($700,000). Ongoing annual operating costs, including platform licenses, data updates, and dedicated staff, are approximately $850,000.

Measured Outcomes

Capital Planning Impact

The most quantifiable outcome has been the restructuring of the county's $8.6 billion five-year Capital Improvement Program. Climate risk scores are now a mandatory input for all infrastructure projects exceeding $5 million. In the first 18 months of operation, the analytics platform influenced prioritization decisions for $1.3 billion in planned infrastructure investments.

Specifically, 14 drainage improvement projects totaling $340 million were re-scoped to account for projected increases in extreme rainfall intensity (20-35% increases in 1-hour, 100-year rainfall by 2050 under SSP2-4.5). Seven of these projects increased design capacities by 25-40%, adding approximately $85 million in costs but extending projected service life by 15-20 years under climate-adjusted conditions.

Three planned road elevation projects in coastal areas were deprioritized after scenario analysis showed that the roads would face chronic tidal flooding (more than 26 days per year) by 2050 regardless of elevation, making managed retreat or alternative access routes more cost-effective over a 30-year planning horizon.

Zoning and Development Decisions

The county integrated climate risk scores into its development review process. All new development applications in areas with First Street Flood Factor scores of 7 or above (out of 10) now require applicants to demonstrate climate-adaptive design features, including elevated mechanical equipment, flood-resistant building materials for the first floor, and on-site stormwater retention exceeding pre-development conditions by 25%.

In the first year, this policy affected 127 development applications covering approximately 3,400 residential units and 890,000 square feet of commercial space. Developers reported average additional construction costs of 3-5% for climate-adaptive features, which were partially offset by lower flood insurance premiums (averaging 15-22% reductions for properties demonstrating adaptive design compliance).

Insurance and Financial Planning

Miami-Dade used the analytics platform to challenge FEMA flood map designations for 12,400 parcels where the parcel-level risk data showed material divergence from FEMA zone classifications. Of these challenges, 3,200 resulted in map amendments that either increased or decreased risk designations. For properties where risk was downgraded, owners saved an estimated $8.7 million annually in National Flood Insurance Program premiums. For properties where risk was upgraded, the county proactively engaged with owners about mitigation options before insurance rate increases took effect.

The county also used scenario modeling to project its own fiscal exposure under different climate trajectories. Under SSP2-4.5, the county estimates cumulative infrastructure damage costs of $3.2-4.8 billion through 2060. Under SSP5-8.5, that range increases to $5.1-7.6 billion. These projections informed the county's decision to increase its catastrophe reserve fund by $120 million over five years and to purchase a parametric insurance product covering extreme rainfall events exceeding 8 inches in 24 hours.

Public Communication

The county launched a public-facing Climate Risk Explorer web application that allows residents to view parcel-level risk scores for flood, heat, and wind hazards. In the first 12 months, the tool received 2.3 million unique page views, with peak traffic coinciding with hurricane season and property transaction periods. Real estate attorneys and title companies in the county reported using the tool during approximately 18% of residential transactions.

What Worked

Parcel-level resolution changed the policy conversation. Previous county risk assessments produced zone-level or neighborhood-level findings that were too coarse to drive specific investment or regulatory decisions. Parcel-level data enabled property-specific policy responses and made climate risk tangible to individual property owners, developers, and elected officials. County commissioners could see exactly which parcels in their districts faced elevated risk, creating political support for adaptive investments.

Dual-vendor approach balanced cost and capability. First Street Foundation's standardized risk scoring provided affordable, comprehensive coverage for all parcels, while Jupiter Intelligence's customized infrastructure analytics delivered the engineering-grade projections needed for capital planning. Using two vendors also created competitive tension that improved responsiveness and data quality.

Validation against historical events built credibility. The county required vendors to demonstrate that their models accurately reproduced observed flooding patterns from Hurricane Irma and Tropical Storm Eta before accepting forward-looking projections. This validation step, which added two months to the timeline, proved essential for gaining trust from engineering staff and elected officials who were initially skeptical of probabilistic climate projections.

What Did Not Work

Scope 3 infrastructure interdependencies remained unmodeled. The analytics platform assessed individual asset risk but could not systematically model cascading failures across interconnected systems (for example, how a flooded electrical substation affects water treatment plant operations). This limitation became apparent during Tropical Storm Nicole in November 2022, when a pump station failure cascaded to affect drainage in areas the model had rated as moderate risk. The county is now investing in network-level resilience modeling to address this gap.

Staff capacity constrained utilization. Despite investing $280,000 in training, the county found that only 35% of eligible planning and engineering staff regularly used the analytics platform 12 months after deployment. The primary barriers were workflow inertia (staff continued using familiar tools and processes), data interpretation uncertainty (staff felt unqualified to explain probabilistic projections to decision-makers), and competing workload demands. The county subsequently hired three dedicated climate analytics specialists to serve as internal consultants for departmental staff.

Private property owner engagement was limited. While the public Climate Risk Explorer generated significant web traffic, the county found that only 12% of property owners in high-risk areas took any documented mitigation action within the first year of receiving risk information. This finding aligns with broader behavioral research showing that risk information alone rarely drives protective action without complementary financial incentives, simplified mitigation pathways, or regulatory requirements.

Climate Risk Analytics KPIs: Benchmark Ranges

Metric	Below Average	Average	Above Average	Top Quartile
Spatial Resolution	Census tract	Neighborhood (250m)	Building footprint	Parcel-level (30m)
Hazard Coverage	Single hazard	2-3 hazards	4-5 hazards	Compound/cascading
Projection Horizon	2050 only	2030-2060	2030-2080 decadal	Annual through 2100
Historical Validation	None	Qualitative	Single event calibration	Multi-event statistical
Staff Utilization Rate	<15%	15-30%	30-50%	>50%
Capital Plan Integration	Ad hoc	Advisory input	Systematic scoring	Mandatory gate criterion
Cost per Parcel Assessed	>$10	$5-10	$2-5	<$2

Transferable Lessons

Other municipalities and utilities considering climate risk analytics deployments can extract several lessons from Miami-Dade's experience:

Start with a specific decision problem, not a general data acquisition exercise. Miami-Dade succeeded because it anchored the project to three defined use cases (capital planning, development review, and insurance management) rather than attempting to create a general-purpose climate intelligence platform.

Budget 40-50% of project costs for data preparation and integration. The climate analytics vendors delivered their products on schedule; the delays and cost overruns occurred in aligning county data systems with the analytics outputs.

Invest in validation before projection. Staff and elected official buy-in depends on demonstrating that the models accurately represent conditions they have personally experienced before asking them to trust forward-looking projections they cannot verify.

Plan for sustained staff capacity, not one-time training. Technology adoption requires dedicated internal champions who can translate analytics outputs into the language and workflows of specific departmental users.

Pair risk information with action pathways. Telling property owners or department heads that risk exists does not change behavior. Providing specific, funded, and simplified response options (such as retrofit incentive programs, pre-approved adaptive design templates, or streamlined permitting for mitigation projects) dramatically increases the likelihood of protective action.

Action Checklist

• Define 2-3 specific decision processes that climate risk analytics will inform before initiating procurement
• Audit existing geospatial, asset management, and financial data systems for completeness and compatibility
• Require vendor proposals to include historical event validation as a deliverable milestone
• Allocate dedicated data engineering resources for integration (not just analytics consumption)
• Establish a cross-departmental climate analytics working group with representatives from planning, engineering, finance, and emergency management
• Build public communication tools that pair risk scores with specific mitigation actions and financial incentives
• Plan for ongoing operating costs of 15-20% of initial implementation annually
• Benchmark against peer jurisdictions using standardized KPIs for resolution, coverage, and staff adoption

Sources

• First Street Foundation. (2025). National Risk Assessment: Methods and Data Documentation, Version 3.0. Brooklyn, NY: First Street Foundation.
• Jupiter Intelligence. (2025). ClimateScore Global: Technical Methodology and Validation Report. San Mateo, CA: Jupiter Intelligence.
• National League of Cities. (2025). Municipal Climate Adaptation Survey: Technology Adoption and Budget Priorities. Washington, DC: NLC.
• Miami-Dade County Office of Resilience. (2025). Climate Risk Analytics Program: Two-Year Implementation Report. Miami, FL: Miami-Dade County.
• NOAA National Centers for Environmental Information. (2025). Billion-Dollar Weather and Climate Disasters: 2020-2025 Summary. Asheville, NC: NCEI.
• Federal Emergency Management Agency. (2025). National Flood Insurance Program: Community Rating System Annual Report. Washington, DC: FEMA.
• Bloomberg Intelligence. (2025). Climate Risk Analytics Market: Sizing, Growth, and Competitive Landscape. New York: Bloomberg LP.