editor_options	markdown wrap 72

(c) Khalil Teber 2024-2025 # Inequality in human development amplifies climate-related disaster risk

Structure of the repository

This repository contains all the scripts necessary to generate all main results and supplementary materials. Scripts used for data preprocessing are also provided to document the full data-generation workflow.

├── README.md
├── analysis                                   <- all scripts used to generate main results and supplementary materials
├── data
│   ├── raw_data                               <- raw data
│   └── intermediate_data                      <- processed data 
│   ├── input                                  <- data used for the analysis
│   └── output                                 <- data produced from analysis
├── figures
│   ├── main figures
│   ├── supplementary_figures
├── preprocess_01_aggregate_climate_variables  <- code needed for aggrgating the climate variables to geocoded events
│   ├── src                         
├── preprocess_02_create_analysis_data         <- code needed for joining disaster, climate and socioeconomic data
│   ├── src                         

Table of Contents

• Climate aggregation
• Create analysis data
• Analysis

Reproducibility

All results presented in the manuscript can be reproduced deterministically using the scripts provided in this repository, together with EM-DAT data obtained directly from the original source.

The recommended reproduction workflow is:

1. Run analysis/0_create_dataframe.R to merge the supplied data with EM-DAT and create the data used for the analysis. output: data/input/disaster_information.csv, the dataframe used for the analysis.

2. Run analysis/0_list_functions.R and analysis/0_plot_settings.R to load needed functions and plot helper functions, themes, color palettes.

3. Run analysis/1_calculate_OR.R to perform the odds ratio analysis. output: csv files with OR calculation and significance testing saved to data/output/odds_sHDI_disasters_df.csv and data/output/odds_disasters_inequalities_df.csv

4. Run analysis/2_generate_main_figures.R to generate the main result figures output: 8 main figures presented in the main text saved to figures/main_figures/...

5. Run analysis/3_generate_supplementary_figures.R to generate the supplementary figures. output: 22 supplementary figures presented in the supplements saved to figures/supplementary_figures/...

The workflow can be executed sequentially using main_script.R, which automates all steps from data integration to figure generation.

To reproduce the Superposed Epoch Analysis (SEA), run the Python script analysis/1_climate_fingerprints.py. The output is saved to the csv files /data/input/disaster_types_hdi_sd_climatology.csv and data/input/disaster_types_sd_climatology.csv. The supplementary tables are created using the notebook: 4_supplementary_tables_S1_S21.Rmd

The preprocessing code is provided to document the full data-generation workflow and ensure methodological transparency. While executing the complete preprocessing pipeline requires substantial computational resources and access to large climate datasets, the workflow is fully specified.

(Preprocessing)-Climate-aggregation

We aggregate the climate variables in three different dimensions (original units, anomalies and zscored anomalies), to obtain time series that summarize the climatological conditions before during and after each event. We use mostly global ERA5 data (0.25deg), and one variable from GLEAM regridded to ERA5 resolution.

(Preprocessing)-Create-analysis-data

In this step, we add the remaining ancillary information needed to the analysis: the subnational human development index (sHDI) and its components, population count and GDP of impacted regions. We keep only the records that have all information essential to the analysis available.

Analysis {#analysis}

In this folder you will find the scripts used to generate all the results of the paper as well as all figures and tables (main results and supplementary materials).

Data

We supply the data needed to generate the results, and some of the data needed for the preprocessing steps. As redistributing EM-DAT data is not allowed, users should download EM-DAT data independently and place the file in data/raw_data, specifying the corresponding file path at the beginning of analysis/0_create_dataframe.R. Administrative boundary geometries in data/input/gdl_map_simplified.gpkg are derived from the Global Data Lab (globaldatalab.org, data accessed August 2025) and are redistributed here for non-commercial use in accordance with GDL terms of use (globaldatalab.org/termsofuse).

Figures

All generated figures will be saved in Figures, main or supplementary_figures subfolders. Code used to generate new figures during peer-review process was added to the corresponding scripts analysis/2_generate_main_figures.R and analysis/3_generate_supplementary_figures.R

Notes

Code added during the peer-review process is integrated directly into the figure-generation scripts.

Session info

To ensure reproducibility, the R session info used to generate the results is provided below.

R version 4.5.2 (2025-10-31) Platform: x86_64-pc-linux-gnu Running under: Ubuntu 24.04.3 LTS

Matrix products: default BLAS: /usr/lib/x86_64-linux-gnu/openblas-pthread/libblas.so.3 LAPACK: /usr/lib/x86_64-linux-gnu/openblas-pthread/libopenblasp-r0.3.26.so; LAPACK version 3.12.0

locale: [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C LC_TIME=en_US.UTF-8 LC_COLLATE=en_US.UTF-8 LC_MONETARY=en_US.UTF-8
[6] LC_MESSAGES=en_US.UTF-8 LC_PAPER=en_US.UTF-8 LC_NAME=C LC_ADDRESS=C LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

time zone: Europe/Berlin tzcode source: system (glibc)

attached base packages: [1] parallel stats graphics grDevices utils datasets methods base

other attached packages: [1] rlang_1.1.6 rnaturalearthdata_1.0.0 rnaturalearth_1.0.1 sf_1.0-21 patchwork_1.3.0 egg_0.4.5
[7] gridExtra_2.3 energy_1.7-12 ggpmisc_0.6.1 ggpp_0.5.9 quantreg_6.1 SparseM_1.84-2
[13] cowplot_1.1.3 xtable_1.8-4 ggpubr_0.6.1 gghighlight_0.5.0 ggforce_0.5.0 ggh4x_0.3.1
[19] ggstream_0.1.0 ggridges_0.5.6 tidytext_0.4.3 lubridate_1.9.4 forcats_1.0.0 stringr_1.5.1
[25] dplyr_1.1.4 purrr_1.0.4 readr_2.1.5 tidyr_1.3.1 tibble_3.3.0 ggplot2_3.5.2
[31] tidyverse_2.0.0