Upcoming Webinar - 3 February 2026

Mapping a Geography of Physical Risk and its Macroeconomic Implications across Different Warming Futures: Introducing EDHEC-CLIRMAP

Tuesday, 3 February 2026
9pm SGT | 1pm GMT | 2pm CET | 8am EST

Explore how climate-change-driven shifts in average temperature could damage regional economic production globally, and discover the quantification work behind EDHEC-CLIRMAP, at the intersection of macroeconomics, climate econometrics, and high-resolution climate data.

In this session, Nicolas Schneider, Senior Research Engineer/Macroeconomist at the EDHEC Climate Institute, will present EDHEC-CLIRMAP, the first interactive online tool offering a high-resolution geographic visualisation of projected climate-induced macroeconomic damages across more than 3,600 subnational regions, covering 95% of global economic production.

The webinar will show:

• Methodology: integration of NASA’s high-resolution climate datasets (CMIP6), subnational economic data, macroeconomic modelling, and climate econometrics.

• Customisation: selecting climate models, warming scenarios (SSP-RCP), and future time horizons.

• Geographic insights: how local damages scale into a global narrative on chronic physical climate risk.

• Applications: financial analysis, public-policy planning, and long-term risk management.

This session is open to investors, consultants, policymakers, researchers, journalists, students, and anyone seeking to understand the macroeconomic implications of climate change.

Register here

INTERVIEW: EDHEC's groundbreaking new tool maps future physical risk and its macro impacts

Over the past 15 years, the quantification (and projection) of the macroeconomic impacts of “physical” climate risk has undergone a revolution. Could you tell us more?

The general, and intuitive, relationship between anthropogenic emissions, rising average temperatures, and the resulting macroeconomic damages stems from highly complex mechanisms. The scientific literature defines damages (Ω) as the percentage deviation of GDP relative to a reference scenario or a historical baseline period. Until roughly 15 years ago, the approach to physical climate risks in macroeconomics was largely global. Knowledge was limited to aggregated predictions, providing little or no insight into the geographical distribution of climate change impacts across continents, countries, or even within countries.

The initial approach developed by Nordhaus (1991) was primarily enumerative, and based on existing literature and sectoral studies. It laid the foundation for a first analytical framework that subsequently informed the DICE models that gave rise to the broader family of Integrated Assessment Models (IAMs)[1].

This approach, however, has faced considerable criticism, notably due to its lack of solid empirical or theoretical foundations (“…damage functions used in most IAMs are completely made up, with no theoretical or empirical foundation,” Pindyck, 2013), as well as certain assumptions deemed excessively conservative. Weitzman (2010) subsequently proposed an alternative, more catastrophic calibration, which also confronted the challenge of rigorous empirical validation.

In 2015, Burke et al. delivered a pioneering study[2], providing new and valuable insights into the distribution of climate-induced damages to GDP across countries. Their main finding intuitively highlights a non-linear relationship between average temperature and GDP growth. It also highlighted the heterogeneity of future damages depending on each country’s starting point (that is, its historical temperature or “baseline”) along this inverted U-shaped curve. However, the methodological developments underlying their work represent a major advance for climate macroeconomics (and macro-econometrics). Their approach to projections is also innovative: based on a single target year (2100) and a given scenario (RCP8.5), they extracted corresponding values for each country from the CMIP5 models, computed a population-weighted average over grid cells, and then linearly backward interpolated temperature values from 2100 to 2010. This method reduced the dimensionality of the projection matrix while providing new geographic information[3].

In 2024, Kotz et al. published one of the first distributions of within-country damages, relying on historical regional economic data (referred to as DOSE[4]) covering 1,661 regions across 88 countries. Their econometric model also presents a significant advance: it is spatial, considering GDP at the regional level, and incorporates a broader range of climate factors (that is, components of physical risk such as extreme precipitation or temperature variability, beyond average temperature alone). Despite this added complexity, the authors confirm that average temperature remains the primary determinant of forecast magnitude. The model also accounts for persistence effects, taking into consideration the influence of past years’ climate on contemporary per capita GDP growth. This represents both a methodological and empirical advance, to the extent that the latest NGFS report (Phase V) by the Network for Greening the Financial System (which is highly valuable for central banks, policymakers, and investors) uses the study’s results to calibrate its chronic physical risk module. Thanks to this approach, the NGFS now provides GDP loss projections distributed by country, year, and scenario, offering a much richer integration of geographic, climatic, and economic information[5].

How does your own work fit into this scientific trajectory?

EDHEC-CLIRMAP fits within this literature. We use more granular raw historical climate data and apply the methodology of Burke et al. (2015) to the within-country geographic resolution introduced by Kotz et al. (2024), before projecting our damage estimates across 3,600 subnational regions representing 95% of global economic output. Moreover, we simplify the visualisation (and therefore the communication) of our findings, while allowing users to explore intuitively the story they convey.

In short, EDHEC-CLIRMAP functions as an interactive world map, where the colour gradient illustrates the uneven geographic distribution of future GDP deviations attributable solely to rising average temperatures (i.e., the central component of chronic physical climate risk[6]) in a warming context. 

Moreover, EDHEC-CLIRMAP provides a methodological foundation[7] for producing geographically disaggregated projections of climate-induced economic damages, consistently covering most administrative regions responsible for 95% of global economic production.

EDHEC-CLIRMAP is a unique proposition. How was this tool developed, and who might be interested in it?

EDHEC-CLIRMAP is, in fact, the culmination of a colossal effort: our own work at EDHEC Business School and the EDHEC Climate Institute, of course, but also that of hundreds of scientists worldwide. We follow an incremental approach (fortunately) building on prior peer-reviewed studies and publicly available data.

EDHEC-CLIRMAP lies at the intersection of several disciplines: macroeconomics, econometrics, and climate science. Its hybrid nature reflects the ongoing multidisciplinary revolution in climate research. Climate change, by its very nature a global problem, cannot be understood solely through standard equilibrium price adjustment mechanisms; it represents a risk to existing economic and sectoral systems and demands a transdisciplinary scientific response. Whether in public research departments or private institutes, work now increasingly combines approaches that previously operated independently (perhaps for the first time) and EDHEC-CLIRMAP is part of this momentum.

 Notes: Projected chronic physical climate damages (%) on per-capita GRP from climate change-driven shifts in average temperature alone, future epoch (e.g., 2021-2040. 2031-2050, ..., 2099) relative to constant historical 1985-2004 temperature means, model means of a selected CMIP6 global climate model (e.g., ACCESS-ESM1-5, BCC·CSM2-MR, ..., UKESM1-0-LL), under a chosen warming scenario(i.e., SSPS-RCP8.5 vigorous warming, SSP2-RCP5 moderate warming, intermediate mid-point scenario between SSPS-8.5 vigorous& SSP2-4.5 moderate warmings).

How was this tool developed?

The execution of this project had to overcome new challenges (at least for economists, though familiar to data scientists) such as the issue of dimensionality: handling, manipulating, and performing operations on extremely large data matrices required specialised programming and the use of high-performance computing systems (supercomputers) to process data of the order of terabytes.

As an example, consider our exercise in projecting macroeconomic damages. This required the prior integration of data from NASA’s NEX-GDDP CMIP6 climate simulations. These datasets provide, for each of the 249,000 unique grid cells covering the Earth’s surface, daily average temperature simulations (365 days), differentiated across 29 climate models (GCMs) and two major climate scenarios (SSP-RCPs), totalling approximately 5.4 billion data points per year. Covering the period from 2025 to 2100, our analytical framework required processing roughly 400 billion data points for the “future” segment of the damage calibration in the EDHEC-CLIRMAP project ^[8].

Why do we do this? The underlying hypothesis (or intuition) – first formulated by Kotz et al. (2024) and adopted here – is that accounting for local variations in climate exposure and within-country economic heterogeneity leads to estimates of aggregate losses that fundamentally challenge the conclusions of earlier studies on the subject.

EDHEC-CLIRMAP therefore allows the user to select:

1. a future time horizon[9] : 2030, 2040, 2050, …, 2099.
2. a global climate model (GCM) from the 29 available options (e.g., ACCESS-ESM1-5, BCC-CSM2-MR, UKESM1-0-LL, etc.).
3. a climate scenario: selected from IPCC temperature pathways, such as SSP5-RCP8.5 for high warming, SSP2-RCP4.5 for moderate warming, or an intermediate synthetic scenario midway between the two trajectories.

Users can then view a global geographic visualisation, where the colour gradient (from blue to red) represents the projected magnitude of damages (%) on regional gross product (RGP) per capita, attributable solely to changes in average temperature (relative to the 1985–2004 historical baseline, i.e., the central component of chronic physical risk), for:

1. the selected future target period;
2. the chosen climate model; and
3. the selected warming scenario (see Figure 2).

Figure 2: Snapshot of EDHEC-CLIRMAP on Provence-Alpes-Cote d’Azur region

Source: https://climateinstitute.edhec.edu/data-visualisations

Who might be interested in EDHEC-CLIRMAP?

We have just launched this tool online, and it is already attracting significant interest: several dozen consultants, investors, asset managers, and fintech professionals have contacted us. However, we believe that EDHEC-CLIRMAP’s audience extends far beyond the financial sector. It can be of interest to the general public, including students, policymakers, local elected officials, researchers, and many others.

EDHEC-CLIRMAP illustrates a projected future, but not an inevitable one: a future that still depends in part on the choices we make today. It also highlights the importance of funding regionally informed adaptation strategies, as the tool identifies areas where net winners and losers, in absolute terms, may emerge.

Anyone can now explore these projections and gain a better understanding of the economic stakes of climate change and its potential macroeconomic implications. Depending on your profile, please feel free to contact us: we provide more detailed data and analyses tailored to different use cases.

References

• Bearpark, T., Hogan, D., & Hsiang, S. (2025). Data anomalies and the economic commitment of climate change. Nature, 644(8075), E7-E11.
• Burke, M., Hsiang, S. M., & Miguel, E. (2015). Global non-linear effect of temperature on economic production. Nature, 527(7577), 235-239.
• Kotz, M., Levermann, A., & Wenz, L. (2024). The economic commitment of climate change. Nature, 628(8008), 551-557.
• Nordhaus, W. D. (1991). To slow or not to slow: the economics of the greenhouse effect. The economic journal, 101(407), 920-937.
• Pindyck, R. S. (2013). Climate change policy: what do the models tell us? Journal of Economic Literature, 51(3), 860-872.
• Schötz, C. (2025). Spatial correlation in economic analysis of climate change. Nature, 644(8076), E27-E30.
• Weitzman, M. L. (2010). What Is The" Damages Function" For Global Warming—And What Difference Might It Make? Climate Change Economics, 1(1), 57-69.
• Wenz, L., Carr, R. D., Kögel, N., Kotz, M., & Kalkuhl, M. (2023). DOSE–Global data set of reported sub-national economic output. Scientific Data, 10(1), 425.