Quantifying Physical Climate Impacts Methodology

PHYSICAL RISK ASSESSMENT & MODELLING Quantifying Climate Impacts: Methodologies for Understanding Physical Risk As physical climate risks transition from abstract future concerns to present-day financial realities, robust methodologies for assessing and quantifying these risks have become essential. Physical climate risk assessment integrates climate science, hazard modeling, exposure analysis, and vulnerability assessment to estimate potential damages and inform adaptation strategies. This section explores the classification of climate hazards, methodological frameworks for risk assessment, modeling approaches, and practical tools enabling financial institutions, corporations, and policymakers to translate climate science into decision-useful risk metrics.

Climate Hazards: Acute and Chronic Classification Physical climate risks stem from two distinct but interrelated categories of hazards, each with different characteristics, time horizons, and financial implications.

Acute Physical Hazards Acute physical hazards are event-driven climate extremes—sudden, discrete occurrences that can cause immediate, severe damage. The Task Force on Climate-related Financial Disclosures (TCFD) defines acute risks as those resulting from "increased severity of extreme weather events." As global temperatures rise, climate science projects both increased frequency and intensified severity of acute hazards.

Key Acute Hazards Tropical Cyclones (Hurricanes, Typhoons, Cyclones): Intense rotating storm systems bringing extreme winds (often exceeding 250 km/h), torrential rainfall, storm surge, and inland flooding. Atlantic hurricane seasons have demonstrated increasing intensity, with 2025 witnessing Category 5 Hurricane Erin—the earliest formation of such intense storms on record.

Flooding: Encompasses multiple flood types with varying drivers: Riverine (Fluvial) Flooding: Overflow of rivers and streams due to extreme precipitation or rapid snowmelt Coastal Flooding: Storm surge, high tides, and sea-level rise inundating coastal areas Pluvial (Surface Water) Flooding: Overwhelmed drainage systems during intense rainfall events Flash Flooding: Sudden, localized flooding from extreme precipitation in short timeframes The Texas flash floods of July 2025, killing 135+ people in Kerr County, exemplify the devastating impact of extreme precipitation events.

Wildfires: Uncontrolled fires in wildlands, forests, or interface areas between natural vegetation and human development. Influenced by temperature, drought, vegetation density, and wind patterns. The 2025 wildfire season saw intensified activity globally, with over USD 80 billion in insured losses from California wildfires alone.

Heatwaves: Extended periods of abnormally high temperatures. Climate change is increasing both frequency and duration of heat extremes. Heatwaves affect labor productivity, strain energy systems, damage crops, and pose direct health threats—particularly to vulnerable populations.

Severe Convective Storms: Thunderstorms producing damaging winds, large hail, and tornadoes. While individual events are localized, aggregate losses from severe convective storms exceeded USD 50 billion in the U.S. in 2024, representing the second-largest source of insured catastrophe losses after hurricanes.

Droughts: While often considered chronic due to multi-year durations, intense short-term droughts can trigger acute impacts—crop failures, water supply crises, wildfire conditions.

Extreme Cold and Freeze Events: Unusually cold temperatures or ice storms. While warming reduces cold extremes in many regions, disruptions to polar vortex can cause severe cold snaps in mid-latitude regions.

Characteristics of Acute Risks Event-Driven: Discrete occurrences with defined start and end dates High Severity: Potential for catastrophic damages over short timeframes Stochastic Nature: Probabilistic rather than deterministic—modeled through return periods and exceedance probabilities Rapid Onset: Limited warning time (hours to days), though forecasting is improving Localized Impact: Geographically specific, though large events can affect entire regions Insurance Relevance: Typically insurable risks, driving catastrophe bond and parametric insurance markets Chronic Physical Hazards Chronic physical hazards involve gradual, long-term shifts in climate patterns. These changes manifest over years to decades, creating sustained stresses on natural and human systems. Chronic hazards are often characterized by shifts in mean conditions or threshold exceedances.

Key Chronic Hazards Rising Mean Temperatures: Gradual increase in average temperatures affects agricultural productivity, energy demand (cooling vs. heating), labor capacity, and ecosystem stability. A 1.5°C increase in global average temperature masks regional variations—some areas experience 3°C+ warming, particularly high latitudes and continental interiors.

Sea-Level Rise: Progressive inundation of coastal areas from thermal expansion of oceans and melting ice sheets/glaciers. Current projections suggest 0.3-1.0 meters of sea-level rise by 2100 under most scenarios, with potential for significantly higher levels if ice sheet dynamics accelerate. Sea-level rise affects coastal real estate, infrastructure, ports, and low-lying island nations.

Changes in Precipitation Patterns: Shifts in rainfall distribution—some regions becoming wetter, others drier. Includes changes in monsoon timing, altered seasonal precipitation, and increased variability. Water stress emerges as critical chronic risk in many regions.

Water Stress and Scarcity: Declining freshwater availability from reduced precipitation, aquifer depletion, glacier retreat, and increased evaporation. Affects agriculture, industrial processes, hydropower generation, and municipal water supplies. The Murray-Darling Basin in Australia and the Colorado River system in the U.S. exemplify water stress driven by chronic climate shifts.

Ocean Acidification: Absorption of atmospheric CO2 by oceans reduces pH, threatening marine ecosystems, particularly coral reefs and shellfish. September 2025 marked the crossing of the seventh planetary boundary when ocean acidification reached critical thresholds.

Permafrost Thawing: Melting of permanently frozen ground in Arctic and high-altitude regions, destabilizing infrastructure, releasing stored carbon and methane, and altering ecosystems.

Ecosystem Degradation and Biodiversity Loss: Chronic temperature and precipitation shifts drive species range migrations, ecosystem composition changes, and in extreme cases, ecosystem collapse. Coral bleaching from sustained elevated ocean temperatures exemplifies chronic ecosystem stress.

Glacial Retreat: Melting of mountain glaciers affects downstream water supplies, particularly in regions dependent on seasonal meltwater (Himalayas, Andes, Alps).

Characteristics of Chronic Risks Gradual Onset: Changes manifest over years to decades rather than days or weeks Persistent: Once established, difficult or impossible to reverse on human timescales Cumulative Impact: Effects accumulate over time, compounding stress on systems Predictable Trends: More amenable to deterministic modeling based on warming trajectories Widespread Geographic Scope: Regional to global impacts rather than localized events Adaptation Imperative: Chronic shifts require proactive adaptation rather than reactive disaster response Compound and Cascading Risks Increasingly, risk assessments must account for compound risks—multiple hazards occurring simultaneously or in sequence—and cascading risks—primary hazards triggering secondary impacts across interconnected systems.

Examples: Drought creating wildfire conditions, with subsequent fires followed by extreme rainfall triggering debris flows on burned watersheds Heatwave increasing electricity demand while reducing power plant efficiency and transmission capacity, causing grid failures Hurricane damage to coastal infrastructure compounded by pre-existing sea-level rise and subsidence Pandemic (like COVID-19) occurring amid climate disasters, straining emergency response capacity The NGFS's 2024 report "Compound Risks: Implications for Physical Climate Scenario Analysis" emphasizes that ignoring compound effects can significantly underestimate actual risk.

Physical Risk Assessment Framework Effective physical climate risk assessment integrates four core components: hazard, exposure, vulnerability, and financial impact transmission. This framework, grounded in disaster risk reduction science and adopted by financial risk management, provides a systematic approach to translating climate science into decision-useful metrics.

Component 1: Hazard Assessment Hazard assessment characterizes the physical phenomena—frequency, intensity, duration, spatial extent, and timing of climate-related events or conditions.

Data Requirements Historical Observations: Weather station data, satellite observations, tide gauges, stream flow records provide baseline understanding of past hazard occurrence. However, historical data alone is insufficient for future risk assessment—climate change is altering hazard distributions, making the past an imperfect guide to the future.

Climate Models: Global Climate Models (GCMs) from the Coupled Model Intercomparison Project (CMIP)— currently Phase 6 (CMIP6)—project future climate under various greenhouse gas concentration pathways (Representative Concentration Pathways or Shared Socioeconomic Pathways). GCMs provide broad climate trends but require downscaling for regional and local analysis.

Catastrophe Models: Probabilistic models simulating tens of thousands of potential events to characterize hazard probability distributions. Commercial vendors (RMS, AIR Worldwide, Moody's) and open-source platforms (CLIMADA, discussed below) provide catastrophe modeling capabilities.

Scenario Selection Physical risk assessment employs multiple climate scenarios reflecting different warming trajectories: Low Warming (SSP1-2.6 or RCP2.6): Assumes aggressive mitigation limiting warming to approximately 1.8°C by 2100. Represents optimistic policy action scenario.

Medium Warming (SSP2-4.5): Mid-range scenario with moderate mitigation, reaching approximately 2.7°C warming by 2100.

High Warming (SSP5-8.5): High emissions scenario with limited mitigation, potentially exceeding 4°C warming by 2100. Often used for stress testing to understand upper-bound risks.

NGFS scenarios (detailed below) provide standardized climate pathways aligned with these SSPs for financial sector use.

Temporal Horizons Physical risk manifests across multiple time horizons: Near-term (2020s-2030s): Relevant for credit risk, capital planning, strategic investments Mid-term (2040s-2050s): Infrastructure investment decisions, long-lived asset valuation Long-term (2060s-2100): Intergenerational equity, sustainable development planning, comprehensive scenario analysis Component 2: Exposure Assessment Exposure measures what is at risk—people, physical assets, economic activities, ecosystems—and their spatial distribution relative to climate hazards.

Asset-Level Granularity Research published in Nature Communications (July 2024) demonstrated that asset-level exposure assessment is critical for accuracy. The study found that investor losses are underestimated by up to 70% when using aggregated rather than asset-level data, and by up to 82% when neglecting acute tail risks. This underscores the importance of granular geolocation of exposed assets.

Exposure Data Sources Corporate Disclosures: Company reports, SEC filings, sustainability reports disclose major facilities, real estate holdings, and operational footprints.

Geospatial Databases: OpenStreetMap, national cadastral databases, commercial real estate databases provide building and infrastructure locations.

Economic Data: GDP distribution, population density, agricultural land use inform macroeconomic exposure assessments.

Supply Chain Mapping: Understanding supplier locations, logistics networks, and critical dependencies reveals indirect exposure through value chain disruption.

Dynamic Exposure Exposure evolves over time due to: Population growth and urbanization shifting concentrations of people and assets Economic development increasing asset values in exposed locations Infrastructure investment creating new exposed capital stock Migration patterns potentially moving populations toward or away from high-hazard areas Component 3: Vulnerability Assessment Vulnerability describes susceptibility to damage when exposed to a hazard of given intensity. A highly exposed asset may suffer minimal damage if resilient, while a highly vulnerable asset may be devastated by moderate hazard intensity.

Vulnerability Functions (Damage Functions) Vulnerability is typically expressed through damage functions relating hazard intensity to degree of damage: Flood Depth-Damage Curves: Relating water depth to percentage damage for different building types Wind Speed-Damage Functions: Hurricane wind speed mapped to structural damage severity Heat-Productivity Curves: Temperature impacts on labor capacity or agricultural yields Fire Intensity-Asset Loss: Wildfire behavior predicting building survival probability Factors Influencing Vulnerability Physical Characteristics: Building materials, structural design, elevation above flood levels, fire-resistant construction Maintenance and Age: Well-maintained, newer infrastructure typically more resilient than aged, degraded assets Adaptive Capacity: Ability to implement protective measures—early warning systems, emergency response capabilities, backup systems Socioeconomic Factors: Wealth, insurance coverage, institutional capacity affect recovery speed and resilience Adaptation Consideration Vulnerability is not static—adaptation investments reduce it. Assessing "residual vulnerability" post-adaptation is crucial for evaluating adaptation effectiveness and identifying remaining risk requiring risk transfer (insurance) or acceptance.

Component 4: Financial Impact Assessment The final component translates physical damages into financial metrics relevant to investors, lenders, insurers, and corporate decision-makers.

Direct Financial Impacts Asset Impairment and Loss: Physical damage to property, plant, equipment requiring repair or replacement Business Interruption: Revenue losses from operational disruptions, supply chain breakdowns, or market access impediments Increased Operating Costs: Higher energy costs from temperature extremes, water scarcity driving up input costs, elevated insurance premiums Stranded Assets: Assets rendered economically unviable due to chronic climate shifts (e.g., coastal real estate with chronic flooding, ski resorts without snow) Indirect Financial Impacts Market Valuation Effects: Stock price reactions to climate disasters, credit rating downgrades for climate exposed entities Demand Shifts: Changing consumer preferences, migration patterns altering market dynamics Liability Risks: Litigation from failure to disclose climate risks or inadequate adaptation Systemic Risks: Correlated losses across portfolios, financial contagion from concentrated climate events Financial Risk Metrics Value-at-Risk (VaR): Maximum potential loss at specified confidence level over given time horizon Expected Loss: Probability-weighted average loss across all modeled scenarios Return Period Losses: Expected damages from 1-in-100 year, 1-in-250 year events (relevant for insurance capital adequacy) Credit Metrics: Probability of default (PD), loss given default (LGD) adjusted for climate risk Portfolio Concentration: Geographic or sectoral clustering of exposure creating correlated risk Modeling Methodologies and Approaches Translating the assessment framework into quantitative risk estimates requires sophisticated modeling methodologies integrating climate science, engineering, economics, and finance.

Probabilistic Catastrophe Modeling Catastrophe models use Monte Carlo simulation to generate synthetic event sets representing tens of thousands of potential disasters. This probabilistic approach captures the full distribution of possible outcomes rather than single point estimates.

Model Structure Hazard Module: Simulates potential events—their location, intensity, frequency. For hurricanes, models generate synthetic storm tracks, wind fields, storm surge, rainfall based on climatological parameters.

Exposure Module: Geo-locates assets, characterizes their value and attributes relevant to vulnerability.

Vulnerability Module: Applies damage functions relating hazard intensity to expected damage for each exposed asset type.

Financial Module: Aggregates individual asset losses, applies insurance terms (deductibles, limits, coverage types), calculates financial metrics.

Climate Change Integration Modern catastrophe models increasingly incorporate climate change through: Adjusting event frequency distributions (more frequent intense hurricanes, heatwaves) Modifying event intensity (higher wind speeds, extreme rainfall) Altering geographic patterns (tropical cyclones reaching higher latitudes) Incorporating chronic baseline shifts (sea-level rise elevating storm surge impacts) Climate Model Downscaling Global Climate Models operate at coarse spatial resolutions (often 100-200 km grid cells), inadequate for local risk assessment. Downscaling bridges this gap through: Statistical Downscaling: Establishing statistical relationships between large-scale GCM outputs and local observations, then applying these relationships to project local futures.

Dynamical Downscaling: Nesting high-resolution regional climate models within GCM boundary conditions, explicitly simulating local atmospheric processes.

Hybrid Approaches: Combining statistical and dynamical methods to optimize accuracy and computational efficiency.

Damage Functions and Impact Modeling The NGFS Phase V scenarios introduced enhanced damage functions for chronic physical risk, calibrated using state-of-the-art climate datasets to capture comprehensive climate impacts beyond simple temperature increases —accounting for precipitation changes, extreme events, and persistence effects on economic productivity.

However, the academic paper (Kotz et al., 2024) underpinning Phase V physical risk estimates has been retracted following peer review critique. Users should interpret Phase V physical damage estimates alongside this uncertainty, while recognizing that NGFS remains best available practice for macroeconomic climate scenarios. The NGFS plans updated scenarios in 2026 incorporating refined methodologies.

Agent-Based and System Dynamics Modeling For complex systems with feedbacks and behavioral responses, agent-based models (simulating individual entities' decisions) and system dynamics models (capturing stocks, flows, and feedback loops) provide insights beyond static damage functions: Simulating migration decisions in response to chronic heat or water stress Modeling supply chain adaptation and reconfiguration Capturing insurance market responses (premium adjustments, coverage withdrawals) Analyzing cascading infrastructure failures NGFS Climate Scenarios: Application for Physical Risk The Network for Greening the Financial System provides authoritative climate scenarios specifically designed for central bank and financial sector use. While detailed NGFS scenario characteristics were covered in the Regulatory Spotlight section, this section emphasizes their practical application for physical risk assessment.

Scenario Narratives and Physical Risk Profiles Orderly Scenarios (Net Zero 2050, Below 2°C) Physical Risk Characteristics: Limited Additional Warming: Warming capped at 1.5-1.8°C by 2100 Contained Acute Hazards: Moderate increases in extreme event frequency/intensity Manageable Chronic Shifts: Sea-level rise constrained to lower projections (30-60 cm by 2100) Adaptation Focus: Proactive investments address emerging risks before they become severe Financial Implications: Lower physical risk losses but higher transition costs from rapid decarbonization. Physical risks still material, requiring adaptation investment, but catastrophic tail risks minimized.

Disorderly Scenarios (Delayed Transition, Divergent Net Zero) Physical Risk Characteristics: Higher Near-Term Emissions: Delayed action allows additional warming in crucial 2020-2040 period Elevated Chronic Baseline: Greater sea-level rise, more pronounced temperature shifts Increased Acute Events: Higher frequency of extreme weather by mid-century Adaptation Catch-Up: Rushed adaptation efforts amid escalating impacts Financial Implications: Compound risks from both elevated physical impacts and abrupt transition policies. Stranded asset risks from delayed transition combined with mounting physical losses.

Hot House World Scenarios (Current Policies, NDCs) Physical Risk Characteristics: Severe Warming: 2.5-4°C+ by 2100, with proportionally larger acute and chronic hazards Catastrophic Tail Risks: Low-probability, high-impact events (ice sheet collapse, ecosystem tipping points) become more likely Chronic Stress Dominance: Persistent heat, water stress, sea-level rise reshape economic geography Adaptation Limits: Some regions approach limits of viable adaptation Financial Implications: Physical risks dominate financial landscape. Economic damages potentially reach 30% of GDP by 2100 (tail risks approaching 50%). Insurance markets face existential challenges in highest-risk regions. Massive capital destruction and reallocation.

Using NGFS Scenarios for Portfolio Analysis Financial institutions apply NGFS scenarios through stress testing exercises:

1. Select Relevant Scenarios: Typically analyze orderly, disorderly, and hot house scenarios to span range of plausible futures

2. Map Portfolio Exposure: Geolocate assets, identify climate-sensitive sectors, map supply chain dependencies

3. Apply Scenario-Specific Hazards: Use NGFS Climate Impact Explorer to extract hazard data (flood extent, temperature anomalies, etc.) for each scenario and time horizon

4. Calculate Scenario-Conditional Losses: Apply vulnerability functions to estimate damages under each scenario

5. Assess Financial Implications: Translate physical losses to credit impacts (PD/LGD changes), market valuations, insurance losses

6. Identify High-Risk Exposures: Detect concentrations, sectors, or geographies with severe scenario conditional losses

7. Inform Risk Management: Use insights to adjust underwriting, capital buffers, investment strategies, engagement priorities

NGFS Short-Term Scenarios for Tactical Planning The May 2025 release of NGFS short-term scenarios (3-5 year horizons) addresses the challenge that long-term pathways have limited relevance for near-term business decisions. Short-term scenarios model: Specific acute event sequences (severe hurricane season, multi-year drought, compound flood-heatwave events) Immediate financial system responses (credit tightening, insurance market disruption) Economic impacts within credit cycles and strategic planning horizons For financial institutions, short-term scenarios enable: Stress testing loan portfolios against plausible near-term disasters Adjusting reinsurance purchasing strategies Informing credit risk models with climate-adjusted default probabilities Scenario planning for business continuity and operational resilience Tools and Platforms for Physical Risk Assessment Translating methodologies into actionable insights requires practical tools. The physical risk assessment ecosystem includes proprietary commercial platforms and open-source solutions.

Commercial Platforms Moody's Climate Solutions (formerly Four Twenty Seven): Provides physical risk scores for individual companies and real estate assets across eight climate hazards, integrating exposure and vulnerability with climate model projections.

S&P Global Sustainable1 Physical Risk Analytics: Asset-based approach covering approximately 21,000 companies with asset-level exposure data. Assesses nine climate hazards (coastal flood, fluvial flood, pluvial flood, extreme heat, extreme cold, tropical cyclone, wildfire, water stress, drought) across four CMIP6 scenarios and decadal time horizons to 2090s.

RMS Climate Change Models: Risk Management Solutions offers catastrophe models incorporating climate change for hurricanes, floods, wildfires, and other perils. Used extensively by insurers and reinsurers for underwriting and capital management.

AIR Climate Models: AIR Worldwide (Verisk) provides probabilistic catastrophe models with climate adjustment factors, enabling scenario-based risk quantification.

Jupiter Intelligence: Climate risk analytics platform using dynamical downscaling to provide high-resolution (neighborhood-level) flood, heat, wind, wildfire, and precipitation projections globally.

Sustainalytics Physical Climate Risk Metrics: Collaboration with XDI (detailed below) providing bottom-up assessment covering 12 million assets across 135 sectors and 235 countries, with metrics on direct and indirect impacts.

XDI (Cross Dependency Initiative): Specializes in asset-level physical climate risk analysis, providing projected financial losses and financial capacity to absorb losses across company portfolios. Underpins Sustainalytics physical risk data.

Open-Source Solutions CLIMADA: Climate Adaptation Platform CLIMADA (CLIMate ADAptation), developed by ETH Zurich's Weather and Climate Risks Group, is the most comprehensive open-source physical climate risk assessment platform. Available in Python, CLIMADA is freely accessible under GNU GPL3 license.

Core Capabilities: Multi-Hazard Modeling: CLIMADA supports tropical cyclones, floods (coastal, riverine, pluvial), droughts, wildfires, earthquakes, winter storms, hail, extreme heat, and more. Hazard modules generate both historical and probabilistic future event sets.

Exposure and Vulnerability Integration: Platform integrates exposure data from multiple sources (OpenStreetMap, national datasets, GDP-based proxies) with vulnerability functions calibrated to specific hazards and asset types.

Probabilistic Risk Quantification: Monte Carlo simulations across event sets generate expected annual loss, return period losses, and full loss distributions—metrics directly comparable to insurance industry standards.

Adaptation Options Appraisal: A distinguishing feature of CLIMADA is systematic assessment of adaptation measures. Users define adaptation interventions (coastal defenses, building retrofits, early warning systems), and CLIMADA calculates damage aversion, costs, and benefit-cost ratios, producing adaptation cost curves ranking measures by efficiency.

Scenario Analysis: CLIMADA integrates climate model projections (CMIP5/CMIP6), enabling assessment under multiple SSP scenarios and time horizons. Users can compare risk today vs. future decades under different warming pathways.

Global Coverage with Local Resolution: CLIMADA operates at consistent 10 km resolution globally, while supporting local applications at 100 m resolution for detailed studies.

Uncertainty and Sensitivity Analysis: Version 3.1.0 introduced comprehensive uncertainty quantification, allowing users to characterize confidence bounds around risk estimates and identify key uncertainty drivers.

Applications: CLIMADA has been applied in over 20 climate adaptation studies worldwide, including: National adaptation planning in Small Island Developing States Insurance sector climate stress testing Infrastructure investment risk assessment Supply chain climate vulnerability analysis International development agency project appraisal User Community and Development: CLIMADA benefits from active community of users and contributors, monthly developers' meetings, extensive documentation, and tutorial resources. The platform's open-source nature enables transparency, reproducibility, and collaborative enhancement—critical for building stakeholder trust in risk assessments.

Accessibility: Available via GitHub with comprehensive documentation at climada.ethz.ch. Installation via Python package managers (conda, pip) with tutorials ranging from introductory (5-12 minutes) to advanced technical applications.

Limitations: As with any model, CLIMADA involves simplifications and uncertainties. Hazard data quality varies by geography (generally better in developed economies). Vulnerability functions require calibration and validation. Climate model uncertainties propagate through analysis. Users must interpret results acknowledging these limitations.

EIOPA CLIMADA-App The European Insurance and Occupational Pensions Authority developed a user-friendly desktop interface for CLIMADA to lower barriers for supervisors and smaller insurers. The CLIMADA-App provides point-and click access to core CLIMADA functionalities without requiring programming skills, though it offers less flexibility than the full Python platform.

Other Open-Source Tools OpenQuake: Global Earthquake Model Foundation's open-source platform for seismic hazard and risk modeling. Particularly relevant for earthquake-exposed portfolios.

Risk Data Hub: European Commission's web platform providing disaster risk assessment data and methodologies covering floods, earthquakes, wildfires, and droughts across Europe.

InaSAFE: Free tool for disaster impact analysis developed by Indonesia's National Disaster Management Agency and World Bank, focusing on flood, earthquake, tsunami, and volcanic hazards.

Key Challenges in Physical Risk Assessment Despite methodological advances, physical climate risk assessment faces persistent challenges: Data Gaps and Quality Geographic Coverage: High-quality hazard data, exposure databases, and vulnerability functions are concentrated in developed economies, particularly North America and Europe. Emerging markets and developing economies often lack granular climate observations, asset databases, and calibrated damage functions.

Supply Chain Transparency: Understanding indirect exposure through value chains requires data on supplier locations and dependencies—information companies often consider proprietary or lack systematically.

Historical Analogs: Climate change is pushing systems into states without historical precedent (unprecedented heat, compound extremes), limiting the relevance of historical damage data for calibrating vulnerability functions.

Model Uncertainty Climate Model Spread: Different GCMs project different regional climate futures even under the same emissions scenario. Ensemble approaches using multiple models mitigate but don't eliminate uncertainty.

Downscaling Accuracy: Statistical and dynamical downscaling introduce additional uncertainties, particularly for extreme events and in topographically complex regions.

Vulnerability Function Precision: Damage functions are often empirically derived from limited observations and may not generalize to future climate conditions or different geographic contexts.

Long Time Horizons vs. Short Decision Cycles Physical risk assessment requires 30-100 year time horizons, while most financial decisions operate on 1-10 year cycles. Reconciling this mismatch requires discounting future climate impacts, but appropriate discount rates for climate risks remain contested.

Non-Stationarity and Tipping Points Climate models generally assume smooth, gradual changes. Abrupt shifts—ice sheet collapse, ocean circulation disruption, permafrost carbon release—could dramatically alter risk profiles but are poorly characterized in current assessments.

Cascading and Compound Risks Most models assess individual hazards independently. Real-world disasters involve compound events and cascading failures across interdependent infrastructure, supply chains, and financial systems. Capturing these interactions remains methodologically challenging.

Adaptation Uncertainty Risk projections depend critically on future adaptation—by households, businesses, infrastructure operators, governments. Modeling adaptation responses involves speculative assumptions about future behavior, technology, investment, and policy.

Best Practices for Physical Risk Assessment Despite challenges, practitioners can enhance assessment quality through established best practices: Use Multiple Scenarios Never rely on single scenario. Analyze range spanning orderly (low warming), disorderly (medium warming), and hot house (high warming) futures to span uncertainty and identify scenario-robust decisions.

Employ Ensemble Approaches Where feasible, use multiple climate models, downscaling techniques, or catastrophe models to characterize model spread and build confidence.

Prioritize Asset-Level Granularity Aggregate, portfolio-level assessments severely underestimate risk. Invest in asset-level geo-location and exposure characterization, as research demonstrates up to 70% underestimation from aggregated approaches.

Integrate Acute and Chronic Risks Assessing only chronic or only acute risks substantially understates total exposure (up to 82% underestimation neglecting acute tail risks). Comprehensive assessment addresses both.

Validate with Historical Events Where possible, backtest models against observed disasters to assess whether damage estimates align with actual outcomes—a reality check on vulnerability functions and exposure data.

Engage Stakeholders Physical risk assessments inform high-stakes decisions affecting communities, workers, and ecosystems. Engaging stakeholders builds trust, incorporates local knowledge improving accuracy, and ensures assessments address relevant questions.

Acknowledge and Communicate Uncertainty Resist false precision. Clearly communicate uncertainty ranges, key assumptions, and model limitations. Decision-makers can manage uncertainty if transparently presented but are misled by spurious precision.

Update Regularly Climate science, hazard data, exposure, and vulnerability evolve. Risk assessments should be updated periodically (every 3-5 years minimum) to incorporate latest climate projections, refined models, and changing exposure.

Connect to Decision-Making Risk assessment is not an end in itself. Ensure outputs directly inform specific decisions—capital allocation, underwriting criteria, adaptation investment prioritization, portfolio construction, disclosure.

The Future of Physical Risk Modeling Physical climate risk assessment is rapidly evolving, with several frontiers promising enhanced capabilities: Machine Learning Integration: AI and machine learning are being applied to improve hazard prediction (nowcasting severe weather), enhance downscaling, detect patterns in complex datasets, and calibrate vulnerability functions from satellite imagery and damage observations.

Real-Time Risk Monitoring: Integration of satellite data, IoT sensors, and continuous climate monitoring enables dynamic risk tracking—updating assessments as hazards evolve rather than static annual assessments.

Supply Chain and Network Modeling: Graph theory and network analysis are being applied to map supply chain vulnerabilities, infrastructure interdependencies, and systemic risk propagation.

Nature-Based Solutions Modeling: Quantifying risk reduction from ecosystem-based adaptation (mangroves for coastal protection, urban forests for heat mitigation) is improving, enabling better cost-benefit analysis of nature-based vs. built adaptation.

Tailored Financial Metrics: Moving beyond generic damage estimates to financial sector-specific outputs— credit scoring models, insurance pricing factors, bank capital calculations, real estate valuations—directly integrating climate risk into core financial processes.

Open Data and Democratization: Initiatives like the NGFS Climate Impact Explorer, CLIMADA, and OpenQuake are democratizing access to risk assessment tools, reducing dependence on proprietary commercial platforms and enabling broader institutional capacity building.

Physical climate risk assessment has matured from academic exercise to core component of financial risk management, corporate strategy, and public policy. While challenges remain—data gaps, model uncertainties, long time horizons—methodologies are advancing rapidly. The combination of standardized frameworks (TCFD, IFRS S2), authoritative scenarios (NGFS), and accessible tools (CLIMADA, commercial platforms) is enabling the financial sector to transition from qualitative acknowledgment of climate risk to quantitative integration into decision-making. The imperative is clear: physical climate risks are here, they are material, and robust assessment methodologies exist to inform resilience-building action.