Climate Impact Map

Map Filters
Display as:
Show me under with a probability
The Climate Impact Lab makes data publicly available under the CC BY 4.0 license.
      Temperature
      Days
      Mortality costs as share of GDP
      Energy expenditures as share of GDP
      View Methodology 

      Updated November 10, 2022

      Climate Projections on Impact Map are based on the World Climate Research Programme’s Sixth Coupled Model Intercomparison Project, known as CMIP6, (O’Neill et al., 2016). Damage (cost) Projections are based on the previous round of scenarios, CMIP5 (Taylor et al., 2012). The Climate Impact Lab’s future research priorities include updating all Damage Projections to be consistent with CMIP6 climate scenarios.

      Climate Projections

      The Climate Projections available here are derived from the Climate Impact Lab’s Global Downscaled Projections for Climate Impacts Research (GDPCIR), freely available on the Microsoft Planetary Computer. The GDPCIR dataset makes use of statistical bias correction and downscaling algorithms, which are specifically designed to accurately represent changes in the extremes. Specifically, the Quantile Delta Mapping (QDM) approach was selected, following the method introduced by Cannon et al. (2015), which preserves quantile-specific trends from global climate models while fitting the full historical distribution for a given day-of-year to a reference dataset (ERA5). To produce GDPCIR, an additional method is introduced, tailored to increase spatial resolution while preserving extreme behavior, Quantile-Preserving Localized-Analog Downscaling (QPLAD). Together, these methods provide a robust means to handle both the central and tail behavior seen in climate model output, while aligning the full historical distribution to a state-of-the-art reanalysis dataset and providing the spatial granularity needed to study surface impacts. A publication providing additional detail is in process and will be linked here as soon as it is available.

      CMIP6 projections do not inherently constitute a probability distribution; rather, they are an ensemble of opportunity, composed of runs conducted by climate modeling teams participating on a voluntary basis and running models that roughly represent ‘best-estimate’ projections of climate behavior. To produce a probabilistic ensemble, we used the Surrogate Model/Mixed Ensemble (SMME) method of Rasmussen et al. (2016). This method weights projections by comparing their global mean surface temperature projections to those of a probabilistic simple climate model, in this case, the Finite Amplified Impulse Response (FaIR) climate model (version 6.2) (Smith et al., 2018). As in Rasmussen et al., potential 2080-2099 global temperature outcomes produced by the probabilistic simple climate model but not represented within the downscaled CMIP6 dataset were represented by ‘model surrogates’, produced using linear pattern scaling of existing global climate models in the GDPCIR, with residuals added to represent high-frequency variability and non-linearities.

      The gridded projections were aggregated to regional estimates by first transforming the daily min, average, or maximum temperature at the grid scale, then aggregating to regions using a weighted average. All variables are weighted using the shares of each region’s population within each grid cell, using gridded population count estimates, adjusted to match 2020 the United Nations World Population Prospects projections (SEDAC, 2020 UN WPP-Adjusted Population Count, v4.11).

      Damage Projections

      Mortality costs

      The methodology for estimating the mortality costs of future climate change is described in full in Carleton et al. (2022). This study uses comprehensive historical mortality records to quantify how death rates across the globe have been affected by observed climate changes.

      Carleton et al. (2020) compile the largest sub-national vital statistics database in the world, detailing 399 million deaths across 41 countries accounting for 55 percent of the global population. By combining these records with decades of detailed daily and local temperature observations, the authors discover that extreme cold and extreme heat have important effects on death rates. These relationships are modified by the climate and income levels of the affected population. Carleton et al. (2020) use these results to model how adaptation affects the sensitivity of a population to extreme temperatures.

      Estimates of the mortality-temperature relationship are used to generate projections of the future impacts of climate change on mortality rates for areas across the globe, dividing the world into 24,378 distinct regions (each containing roughly 300,000 people, about the size of a U.S. county). Using a revealed preference technique to measure the total cost of adaptive behaviors and technologies, these projections capture the full mortality risk of climate change, accounting for both adaptation benefits and costs, in addition to direct mortality impacts.

      These estimates are based upon emissions scenario RCP 4.5 or RCP 8.5, socioeconomic scenario SSP3 (from the IIASA Shared Socioeconomic Pathways database), and are climate model-weighted means over 33 climate models and 1,000 Monte Carlo simulation runs, allowing for an assessment of the uncertainty surrounding any particular projection. The full estimates also reflect statistical uncertainty related to the underlying economic and health data.

      Projected impacts of climate change on mortality rates are then monetized and used to determine the costs of excess mortality risk in a given year. The full mortality risk of climate change mapped here includes the monetized value of both changes in mortality rates and changes in expenditures on adaptation. Damages are valued at an income-scaled value of statistical life (VSL) relying the U.S. EPA’s VSL estimate of $10.95 million (2019 USD). Damages aggregated at a higher geographical level than impact region are totals of the corresponding impact region-level estimates (there are 24,378 impact regions across the globe). Damages are presented as a percent change of projected Gross Domestic Product in each period, based upon socioeconomic scenario SSP3 (from the IIASA Shared Socioeconomic Pathways database).

      Energy costs

      The methodology for estimating the energy use costs of future climate change is described in full in Rode et al. (2021). This study uses comprehensive historical energy consumption data derived from International Energy Agency data files to quantify how a population’s use of electricity and other fuels (for example, natural gas, oil and coal) energy consumption responds to climate changes. The authors utilize the World Energy Balances dataset of the International Energy Agency, which describes electricity and direct fuel consumption across residential, commercial, industrial, and agricultural end-uses in 146 countries during 1971-2010.

      By combining these records with decades of detailed daily and local temperature observations, the authors discover that extreme cold and extreme heat have important effects on energy consumption. These relationships differ by energy type (electricity, other fuels) and are modified by the income levels and climate of the affected population. The study uses these results to model how income growth and adaptation affect the sensitivity of energy consumption to extreme temperatures.

      The authors then use these estimates of the energy-temperature relationship to generate projections of the future impacts of climate change on electricity and direct fuel consumption for areas across the globe, dividing the world into 24,378 distinct regions. Each region contains roughly 300,000 people—about the size of a U.S. county. The projected impacts capture the effects of adaptive behaviors that populations undertake as they become richer and exposed to warmer climates.

      These estimates are based upon emissions scenario RCP 4.5 or RCP 8.5, socioeconomic scenario SSP3 (from the IIASA Shared Socioeconomic Pathways database), and are climate model-weighted means over 33 climate models and 1,000 Monte Carlo simulation runs, allowing for an assessment of the uncertainty surrounding any particular projection. The full estimates also reflect statistical uncertainty related to the underlying data.

      Projected changes in energy expenditures (both electricity and other fuels) are then monetized and used to determine the costs in a given year. Quantiles are calculated using the delta method along with Newton’s method. Expenditure changes and GDP used to construct percentages are aggregated to higher geographical levels as totals of the corresponding impact region-level estimates (there are 24,378 impact regions across the globe). Future prices are assumed to grow at 1.4% per year. Damages are presented as a percent change of projected Gross Domestic Product in each period, based upon socioeconomic scenario SSP3.

      References

      Cannon, Alex J., Stephen R. Sobie, and Trevor Q. Murdock. “Bias Correction of GCM Precipitation by Quantile Mapping: How Well Do Methods Preserve Changes in Quantiles and Extremes?”, Journal of Climate 28, 17 (2015): 6938-6959, accessed Nov 10, 2022, https://doi.org/10.1175/JCLI-D-14-00754.1

      Center for International Earth Science Information Network – CIESIN – Columbia University. “Gridded Population of the World, Version 4 (GPWv4): Population Count Adjusted to Match 2015 Revision of UN WPP Country Totals, Revision 11.” Palisades, New York: NASA Socioeconomic Data and Applications Center (SEDAC), accessed Nov 10, 2022, https://doi.org/10.7927/H4PN93PB.

      Carleton, T., Jina, A., Delgado, M., et al. “Valuing the Global Mortality Consequences of Climate Change Accounting for Adaptation Costs and Benefits”, The Quarterly Journal of Economics, 2022;, qjac020, https://doi.org/10.1093/qje/qjac020

      O’Neill, B. C., Tebaldi, C., van Vuuren, D. P., et al. “The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6”, Geosci. Model Dev., 9, (2016): 3461–3482, accessed Nov 10, 2022, https://doi.org/10.5194/gmd-9-3461-2016.

      Rasmussen, D. J., Malte Meinshausen, and Robert E. Kopp. “Probability-Weighted Ensembles of U.S. County-Level Climate Projections for Climate Risk Analysis”, Journal of Applied Meteorology and Climatology 55, 10 (2016): 2301-2322, accessed Nov 10, 2022, https://doi.org/10.1175/JAMC-D-15-0302.1

      Rode, A., Carleton, T., Delgado, M., et al. “Estimating a social cost of carbon for global energy consumption”, Nature (2021). https://doi.org/10.1038/s41586-021-03883-8

      Smith, C. J., Forster, P. M., Allen, M., Leach, N., Millar, R. J., Passerello, G. A., and Regayre, L. A.: FAIR v1.3: a simple emissions-based impulse response and carbon cycle model, Geosci. Model Dev., 11, 2273–2297, https://doi.org/10.5194/gmd-11-2273-2018, 2018.

      Taylor, Karl E., Ronald J. Stouffer, and Gerald A. Meehl. “An Overview of CMIP5 and the Experiment Design”, Bulletin of the American Meteorological Society 93, 4 (2012): 485-498, accessed Nov 10, 2022, https://doi.org/10.1175/BAMS-D-11-00094.1

      Download Data

      Climate Impact Map

      Map Filters
      Display as:
      Show Me under with a probability
      The Climate Impact Lab makes data publicly available under the CC BY 4.0 license.
          Temperature
          Days
          Mortality costs as share of GDP
          Energy expenditures as share of GDP
          View Methodology 

          Updated November 10, 2022

          Climate Projections on Impact Map are based on the World Climate Research Programme’s Sixth Coupled Model Intercomparison Project, known as CMIP6, (O’Neill et al., 2016). Damage (cost) Projections are based on the previous round of scenarios, CMIP5 (Taylor et al., 2012). The Climate Impact Lab’s future research priorities include updating all Damage Projections to be consistent with CMIP6 climate scenarios.

          Climate Projections

          The Climate Projections available here are derived from the Climate Impact Lab’s Global Downscaled Projections for Climate Impacts Research (GDPCIR), freely available on the Microsoft Planetary Computer. The GDPCIR dataset makes use of statistical bias correction and downscaling algorithms, which are specifically designed to accurately represent changes in the extremes. Specifically, the Quantile Delta Mapping (QDM) approach was selected, following the method introduced by Cannon et al. (2015), which preserves quantile-specific trends from global climate models while fitting the full historical distribution for a given day-of-year to a reference dataset (ERA5). To produce GDPCIR, an additional method is introduced, tailored to increase spatial resolution while preserving extreme behavior, Quantile-Preserving Localized-Analog Downscaling (QPLAD). Together, these methods provide a robust means to handle both the central and tail behavior seen in climate model output, while aligning the full historical distribution to a state-of-the-art reanalysis dataset and providing the spatial granularity needed to study surface impacts. A publication providing additional detail is in process and will be linked here as soon as it is available.

          CMIP6 projections do not inherently constitute a probability distribution; rather, they are an ensemble of opportunity, composed of runs conducted by climate modeling teams participating on a voluntary basis and running models that roughly represent ‘best-estimate’ projections of climate behavior. To produce a probabilistic ensemble, we used the Surrogate Model/Mixed Ensemble (SMME) method of Rasmussen et al. (2016). This method weights projections by comparing their global mean surface temperature projections to those of a probabilistic simple climate model, in this case, the Finite Amplified Impulse Response (FaIR) climate model (version 1.6.2) (Smith et al., 2018). As in Rasmussen et al., potential 2080-2099 global temperature outcomes produced by the probabilistic simple climate model but not represented within the downscaled CMIP6 dataset were represented by ‘model surrogates’, produced using linear pattern scaling of existing global climate models in the GDPCIR, with residuals added to represent high-frequency variability and non-linearities.

          The gridded projections were aggregated to regional estimates by first transforming the daily min, average, or maximum temperature at the grid scale, then aggregating to regions using a weighted average. All variables are weighted using the shares of each region’s population within each grid cell, using gridded population count estimates, adjusted to match 2020 the United Nations World Population Prospects projections (SEDAC, 2020 UN WPP-Adjusted Population Count, v4.11).

          Damage Projections

          Mortality costs

          The methodology for estimating the mortality costs of future climate change is described in full in Carleton et al. (2022). This study uses comprehensive historical mortality records to quantify how death rates across the globe have been affected by observed climate changes.

          Carleton et al. (2020) compile the largest sub-national vital statistics database in the world, detailing 399 million deaths across 41 countries accounting for 55 percent of the global population. By combining these records with decades of detailed daily and local temperature observations, the authors discover that extreme cold and extreme heat have important effects on death rates. These relationships are modified by the climate and income levels of the affected population. Carleton et al. (2020) use these results to model how adaptation affects the sensitivity of a population to extreme temperatures.

          Estimates of the mortality-temperature relationship are used to generate projections of the future impacts of climate change on mortality rates for areas across the globe, dividing the world into 24,378 distinct regions (each containing roughly 300,000 people, about the size of a U.S. county). Using a revealed preference technique to measure the total cost of adaptive behaviors and technologies, these projections capture the full mortality risk of climate change, accounting for both adaptation benefits and costs, in addition to direct mortality impacts.

          These estimates are based upon emissions scenario RCP 4.5 or RCP 8.5, socioeconomic scenario SSP3 (from the IIASA Shared Socioeconomic Pathways database), and are climate model-weighted means over 33 climate models and 1,000 Monte Carlo simulation runs, allowing for an assessment of the uncertainty surrounding any particular projection. The full estimates also reflect statistical uncertainty related to the underlying economic and health data.

          Projected impacts of climate change on mortality rates are then monetized and used to determine the costs of excess mortality risk in a given year. The full mortality risk of climate change mapped here includes the monetized value of both changes in mortality rates and changes in expenditures on adaptation. Damages are valued at an income-scaled value of statistical life (VSL) relying the U.S. EPA’s VSL estimate of $10.95 million (2019 USD). Damages aggregated at a higher geographical level than impact region are totals of the corresponding impact region-level estimates (there are 24,378 impact regions across the globe). Damages are presented as a percent change of projected Gross Domestic Product in each period, based upon socioeconomic scenario SSP3 (from the IIASA Shared Socioeconomic Pathways database).

          Energy costs

          The methodology for estimating the energy use costs of future climate change is described in full in Rode et al. (2021). This study uses comprehensive historical energy consumption data derived from International Energy Agency data files to quantify how a population’s use of electricity and other fuels (for example, natural gas, oil and coal) energy consumption responds to climate changes. The authors utilize the World Energy Balances dataset of the International Energy Agency, which describes electricity and direct fuel consumption across residential, commercial, industrial, and agricultural end-uses in 146 countries during 1971-2010.

          By combining these records with decades of detailed daily and local temperature observations, the authors discover that extreme cold and extreme heat have important effects on energy consumption. These relationships differ by energy type (electricity, other fuels) and are modified by the income levels and climate of the affected population. The study uses these results to model how income growth and adaptation affect the sensitivity of energy consumption to extreme temperatures.

          The authors then use these estimates of the energy-temperature relationship to generate projections of the future impacts of climate change on electricity and direct fuel consumption for areas across the globe, dividing the world into 24,378 distinct regions. Each region contains roughly 300,000 people—about the size of a U.S. county. The projected impacts capture the effects of adaptive behaviors that populations undertake as they become richer and exposed to warmer climates.

          These estimates are based upon emissions scenario RCP 4.5 or RCP 8.5, socioeconomic scenario SSP3 (from the IIASA Shared Socioeconomic Pathways database), and are climate model-weighted means over 33 climate models and 1,000 Monte Carlo simulation runs, allowing for an assessment of the uncertainty surrounding any particular projection. The full estimates also reflect statistical uncertainty related to the underlying data.

          Projected changes in energy expenditures (both electricity and other fuels) are then monetized and used to determine the costs in a given year. Quantiles are calculated using the delta method along with Newton’s method. Expenditure changes and GDP used to construct percentages are aggregated to higher geographical levels as totals of the corresponding impact region-level estimates (there are 24,378 impact regions across the globe). Future prices are assumed to grow at 1.4% per year. Damages are presented as a percent change of projected Gross Domestic Product in each period, based upon socioeconomic scenario SSP3.

          References

          Cannon, Alex J., Stephen R. Sobie, and Trevor Q. Murdock. “Bias Correction of GCM Precipitation by Quantile Mapping: How Well Do Methods Preserve Changes in Quantiles and Extremes?”, Journal of Climate 28, 17 (2015): 6938-6959, accessed Nov 10, 2022, https://doi.org/10.1175/JCLI-D-14-00754.1

          Center for International Earth Science Information Network – CIESIN – Columbia University. “Gridded Population of the World, Version 4 (GPWv4): Population Count Adjusted to Match 2015 Revision of UN WPP Country Totals, Revision 11.” Palisades, New York: NASA Socioeconomic Data and Applications Center (SEDAC), accessed Nov 10, 2022, https://doi.org/10.7927/H4PN93PB.

          Carleton, T., Jina, A., Delgado, M., et al. “Valuing the Global Mortality Consequences of Climate Change Accounting for Adaptation Costs and Benefits”, The Quarterly Journal of Economics, 2022;, qjac020, https://doi.org/10.1093/qje/qjac020

          O’Neill, B. C., Tebaldi, C., van Vuuren, D. P., et al. “The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6”, Geosci. Model Dev., 9, (2016): 3461–3482, accessed Nov 10, 2022, https://doi.org/10.5194/gmd-9-3461-2016.

          Rasmussen, D. J., Malte Meinshausen, and Robert E. Kopp. “Probability-Weighted Ensembles of U.S. County-Level Climate Projections for Climate Risk Analysis”, Journal of Applied Meteorology and Climatology 55, 10 (2016): 2301-2322, accessed Nov 10, 2022, https://doi.org/10.1175/JAMC-D-15-0302.1

          Rode, A., Carleton, T., Delgado, M., et al. “Estimating a social cost of carbon for global energy consumption”, Nature (2021). https://doi.org/10.1038/s41586-021-03883-8

          Smith, C. J., Forster, P. M., Allen, M., Leach, N., Millar, R. J., Passerello, G. A., and Regayre, L. A.: FAIR v1.3: a simple emissions-based impulse response and carbon cycle model, Geosci. Model Dev., 11, 2273–2297, https://doi.org/10.5194/gmd-11-2273-2018, 2018.

          Taylor, Karl E., Ronald J. Stouffer, and Gerald A. Meehl. “An Overview of CMIP5 and the Experiment Design”, Bulletin of the American Meteorological Society 93, 4 (2012): 485-498, accessed Nov 10, 2022, https://doi.org/10.1175/BAMS-D-11-00094.1

          Download Data