Fully replicable code performing the health impact projections reported in the paper:
Masselot, P., Mistry, M.N., Rao, S. et al. Estimating future heat-related and cold-related mortality under climate change, demographic and adaptation scenarios in 854 European cities. Nat Med (2025). https://doi.org/10.1038/s41591-024-03452-2
Results from the paper can be explroed interactively in a dedicated Shiny app.
Warning
Reproducing the full results in extremely intensive computationally and memory wise. It is recommended to test the scripts on a reduced number of cities, and with a reduced number of simulations (see the nsim parameter). Also consider changing the parameters ncores and grpsize that also control the speed and memory efficiency of the analysis. See script 01_pkg_params.R
The input data necessary to run the analysis are stored on a dedicated Zenodo repository. Download the data.zip archive and extract the files into a data folder in the project directory. this can be performed directly from R (see the first script 01_pkg_params.R).
Tables and plots can be replicated without running the analysis by downloading and extracting the files in results_parquet found on the Zenodo repository. See additional details below
The scripts are run in order. Scripts 01 to 03 run the main analysis with results saved at the end of 03_attribution.R. Scripts 04 to 06 can then be ran by loading results saved beforehand.
| Script | Descriptions |
|---|---|
01_pkg_params.R |
Load the necessary R libraries and defines all the analysis parameters. Also contains some parameters related to parallelization of the main loop which should be considered carefully before running a full analysis |
02_prep_data.R |
Load and prepare data about cities, demographic projections, and global warming levels |
03_attribution.R |
Centrepiece of the analysis. Loops through cities and scenarios to perform the health impact projections. For each city, the code loads temperature series and exposure-response function data, and compute the health impact of temperature. Health impacts are then summarised and saved in a temporary folder. Results are then aggregated once the health impact for all the cities from a country or regions have been computed. Note that the loop is stratified into chunks to avoid storing too much intermediary data simulataneously. |
04_tables.R |
Produces the main table from the article. Note that this script will load previously saved results and can then be used without having to run the previous scripts. |
05_plots.R |
Produces plots featured in the main text of the article. |
06_plot_supp.R |
Produces Extended Data plots. |