Climate Data Sources for ESG Investors
Where Do Investors Find Reliable Climate Data?
Climate analysis requires data — emissions totals, energy consumption, physical asset locations, scenario pathways — and the quality of that data varies enormously across sources. Raw corporate disclosures, regulatory filings, modeled estimates, satellite observations, and proprietary commercial datasets each have different strengths, limitations, and coverage gaps. Understanding the data supply chain is essential for any investor who wants to build climate analysis that holds up to scrutiny.
Climate data sources are the databases, platforms, and reporting frameworks from which investors obtain the emissions, energy, physical risk, and scenario data needed to assess portfolio climate exposure and alignment.
Key Takeaways
- No single source covers all companies comprehensively; investors typically layer multiple datasets.
- CDP is the primary source for voluntary corporate disclosure; EDGAR and EPA provide regulatory-grade country and facility-level data.
- The IEA provides authoritative energy statistics and net-zero scenario pathways used in transition risk analysis.
- Commercial ESG data providers (MSCI, Sustainalytics, Bloomberg, Refinitiv) combine reported data with modeled estimates for universe-wide coverage.
- Data quality varies significantly: disclosed Scope 3 data is especially unreliable; third-party-verified data is more trustworthy than unverified self-reports.
The Data Landscape
Climate data for investors can be organized into five categories: voluntary corporate disclosure, regulatory emissions registries, energy and scenario data, physical risk datasets, and commercial aggregation services. Each category serves different analytical needs.
Why Data Quality Matters
Investors who rely on low-quality data risk misallocating capital — overestimating the climate performance of high-emitting companies that under-report, underestimating transition risk for companies that do not disclose at all. As regulatory requirements expand (ISSB S2, CSRD, SEC climate rules), disclosed data quality is improving, but gaps persist, particularly for smaller companies, emerging-market issuers, and Scope 3 categories.
Voluntary Corporate Disclosure
CDP (formerly Carbon Disclosure Project)
CDP operates the world's largest voluntary climate disclosure system. Over 23,000 companies responded to CDP questionnaires in 2023, representing more than half of global market capitalization. CDP collects detailed Scope 1, 2, and 3 emissions data, energy consumption, climate targets, governance structures, and risk assessments through standardized questionnaires aligned with TCFD.
CDP scores companies from A (leadership) to D- (disclosure) on climate, water, and forests. The A List, published annually, identifies companies demonstrating best practice. CDP data is freely available to investors on the CDP Data Portal (investor access requires registration and agreement to terms). Key limitations: coverage is concentrated in large-cap, developed-market companies; response quality varies; Scope 3 data from CDP questionnaires is self-reported without mandatory third-party assurance.
GRI Universal Standards
The Global Reporting Initiative standards — particularly GRI 305 (Emissions), GRI 302 (Energy), and GRI 303 (Water and Effluents) — provide the most widely adopted framework for sustainability reports. GRI data is embedded in company sustainability reports and, increasingly, structured in machine-readable XBRL format under CSRD. GRI data is useful for company-level deep dives but requires manual extraction from PDF reports at scale.
TCFD Reports
Since 2017, thousands of companies have published TCFD-aligned disclosures. In 2023, TCFD was formally absorbed into the ISSB framework (IFRS S2), which inherits and extends TCFD's four-pillar structure. TCFD/ISSB disclosures include scenario analysis results, physical risk assessments, and transition risk descriptions that are not available in pure emissions datasets.
Regulatory Emissions Registries
EDGAR (Emissions Database for Global Atmospheric Research)
EDGAR, maintained by the European Commission's Joint Research Centre and the International Institute for Applied Systems Analysis (IIASA), provides global greenhouse gas emission inventories for all countries from 1970 onward. EDGAR covers CO₂, CH₄, N₂O, and fluorinated gases at country and sector level, derived from energy statistics, industrial production data, and agricultural activity.
EDGAR is the definitive source for country-level historical emissions analysis and is used by the IPCC as one of its primary inputs. It is publicly available at edgar.jrc.ec.europa.eu. EDGAR does not provide company-level data — it operates at country and sector granularity.
US EPA Greenhouse Gas Reporting Program (GHGRP)
The US EPA GHGRP requires annual Scope 1 emissions reporting from US facilities emitting 25,000 metric tons of CO₂-equivalent or more per year. Over 8,000 US facilities report annually, covering roughly 85–90% of US industrial GHG emissions. Data is publicly available through EPA's Facility Level Information on GHGs Tool (FLIGHT). GHGRP data is facility-specific and verified, making it the most reliable source for US industrial emitters.
EU Emissions Trading System (EU ETS)
The European Union Emissions Trading System requires annual verified emissions reporting from all installations covered by the ETS — approximately 10,000 power plants, industrial facilities, and airlines. EU ETS data is published by the European Environment Agency and is facility-specific, third-party verified, and publicly accessible. For large European industrial companies and utilities, EU ETS data often provides the most accurate available emissions figures.
UK and Other National Registries
The UK operates its own ETS following Brexit. Australia, New Zealand, Canada, South Korea, Japan, and China operate national carbon registries with varying coverage and public disclosure requirements. Coverage outside the EU and US is more fragmented and less consistently public.
Energy and Scenario Data
IEA (International Energy Agency)
The IEA publishes the most authoritative global energy statistics and decarbonization scenario analysis. Key products include:
- World Energy Outlook (WEO): Annual publication with three scenarios — Announced Pledges Scenario (APS), Stated Policies Scenario (STEPS), and Net Zero Emissions by 2050 (NZE) — providing technology deployment pathways, fuel demand projections, and investment requirements.
- Energy Technology Perspectives (ETP): Deeper technology-level analysis of clean energy transition pathways.
- Monthly Electricity Statistics: Real-time renewable electricity generation data by country.
- Tracking Clean Energy Progress: Annual assessment of whether clean energy technologies are on track for the NZE scenario.
The IEA's NZE scenario, published in 2021, is the primary benchmark used by the IPCC, SBTi, and most transition risk models for aligning investment portfolios. It provides sector-specific decarbonization pathways that PACTA and TPI use as reference scenarios. IEA data is available at iea.org; some premium datasets require paid access.
NGFS Climate Scenarios
The Network for Greening the Financial System, a coalition of central banks and supervisors, publishes standardized climate scenarios used in financial stress testing and portfolio risk analysis. NGFS scenarios span four categories: orderly transition (Early Action, Net Zero 2050), disorderly transition (Delayed Transition, Divergent Net Zero), hot house world, and below 2°C. These scenarios underpin most bank and insurer climate stress tests, including the Bank of England Climate Biennial Exploratory Scenario (CBES) and ECB climate stress tests.
NGFS scenario data is freely available at ngfs.net and through the Climate Impact Explorer tool maintained by Climate Analytics. Physical damage functions and GDP impact estimates from the NGFS scenarios are used in climate VaR models.
Physical Risk Datasets
Four Twenty Seven (Moody's)
Four Twenty Seven provides physical climate risk scores for over 2,500 listed companies and millions of assets worldwide. Risk scores cover flood, heat stress, hurricane/typhoon, sea level rise, water stress, and wildfire, derived from CMIP6 climate model projections under RCP 2.6, 4.5, and 8.5 pathways. Moody's acquired Four Twenty Seven in 2019; data is available through the Moody's ESG Solutions platform.
Jupiter Intelligence (S&P Global)
Jupiter Intelligence provides property-level physical risk scores and financial loss estimates for real estate, infrastructure, and corporate assets. S&P Global acquired Jupiter in 2021; data is embedded in the S&P Global Sustainable1 platform.
Verisk (Extreme Events Solutions)
Verisk's property analytics capabilities cover catastrophe modeling for hurricanes, flood, wildfire, and winter storms, with applications to insurance portfolios and corporate real estate risk assessment.
Climate Impact Explorer
A free tool maintained by Climate Analytics that visualizes NGFS and IPCC physical climate projections at country, city, and watershed level for non-specialist users.
Commercial ESG Data Aggregators
MSCI ESG Research
MSCI provides ESG ratings, carbon footprint data, climate VaR estimates, temperature alignment scores (ITR), and physical risk scores for over 14,000 listed companies. MSCI combines CDP and other voluntary disclosures with proprietary models for companies that do not disclose. MSCI's financed emissions calculation follows the PCAF standard. Data is available through the MSCI Climate Lab platform.
Sustainalytics (Morningstar)
Sustainalytics covers over 20,000 companies with ESG risk ratings, carbon metrics, and UNGC compliance scores. Sustainalytics's Low Carbon Transition Rating (LCTR) assesses companies' exposure to and management of carbon transition risk on a 0–100 scale. Morningstar acquired Sustainalytics in 2020; data is available through the Morningstar Direct platform.
Bloomberg ESG Data
Bloomberg embeds ESG data including emissions, energy consumption, water, waste, and governance metrics directly in the Bloomberg Terminal (field mnemonics ESG_DISCLOSURE_SCORE, GHG_SCOPE_1, GHG_SCOPE_2, etc.). Bloomberg sources primarily company-reported data without significant modeling infill, meaning coverage gaps are visible as missing data rather than modeled estimates — an advantage for investors who prefer transparency about data provenance.
Refinitiv (LSEG)
Refinitiv ESG scores cover over 12,000 companies across 400+ environmental, social, and governance measures, combining company disclosures with alternative data sources. Refinitiv is now part of LSEG (London Stock Exchange Group).
ISS ESG
Institutional Shareholder Services ESG data covers climate, governance, and social metrics with particular strength in proxy voting analytics and ESG in governance research.
Data Quality and Verification
The Modeled Estimate Problem
For companies that do not disclose, commercial providers generate modeled estimates based on sector, revenue, geography, and peer comparisons. Modeled estimates can introduce systematic bias — typically underestimating emissions for sectors with poor voluntary disclosure, creating an artificial "halo" for non-disclosers. Investors should track what percentage of portfolio emissions data is reported versus modeled, and apply higher uncertainty ranges to modeled figures.
Third-Party Assurance
CDP, GRI, and ISSB all encourage but do not universally require third-party assurance of emissions data. CSRD mandates limited assurance for all reported ESRS data (with a pathway to reasonable assurance), significantly improving European data quality post-2025. SFDR PAI reporting does not yet require assurance, though ESMA has proposed strengthening this.
Scope 3 Reliability
Scope 3 data is the least reliable category across all sources. CDP Scope 3 responses show wide variation in methodology, completeness, and category coverage. Until supply-chain data platforms (Supplier.io, EcoVadis, Sphera) achieve broad supplier coverage and standardized methodology, Scope 3 should be treated as directional rather than precise. The VCMI (Voluntary Carbon Markets Integrity initiative) and SBTi both acknowledge these limitations in their Scope 3 target-setting guidance.
Common Mistakes
Using a single data source for portfolio-wide analysis. Each source has coverage gaps; combining CDP reported data with modeled infill from commercial providers and regulatory data where available produces more complete results.
Treating modeled estimates as equally reliable as reported data. Always tag data fields with source type (reported/estimated/modeled) and apply higher analytical tolerance to modeled figures.
Ignoring base year vintage. Companies change their base years for emissions intensity targets; comparing companies with different base year vintages without adjustment produces misleading benchmark conclusions.
Overlooking currency of data. Climate disclosure lags by 12–18 months; a 2024 Bloomberg terminal query typically returns 2022 or 2023 reported data. This matters for rapidly decarbonizing sectors like European utilities.
Frequently Asked Questions
Is CDP data free to use? CDP provides free access to its dataset for investors who sign up on the CDP Data Portal and agree to its terms of use. Some advanced analytical tools and company-level reports require institutional subscriptions.
How does EDGAR differ from the EPA GHGRP? EDGAR is a country-level academic dataset derived from economic activity statistics; EPA GHGRP is a regulatory system with facility-level mandatory reporting and verification. For US industrial facilities, GHGRP is more granular and reliable. EDGAR is essential for country-level and sector-level global analysis.
What is the PCAF standard? The Partnership for Carbon Accounting Financials (PCAF) published the Global GHG Accounting and Reporting Standard for the Financial Industry in 2020. It provides methodology for banks and investors to calculate financed emissions (Scope 3 Category 15) across loans, bonds, equity, real estate, and other asset classes. PCAF scores data quality on a scale of 1 (verified reported) to 5 (estimated without data), improving transparency about data reliability.
Related Concepts
Summary
Climate data for investors is abundant but uneven. Voluntary disclosure through CDP and GRI, regulatory registries through EPA GHGRP and EU ETS, energy and scenario data through the IEA and NGFS, physical risk data through providers such as Four Twenty Seven and Jupiter, and commercial aggregation through MSCI, Sustainalytics, and Bloomberg all contribute complementary pieces. Quality declines as coverage expands: verified facility data is most reliable; modeled Scope 3 estimates are least reliable. Building analytical processes that clearly distinguish data provenance, flag uncertainty, and weight conclusions accordingly is the defining discipline of serious climate investment research.