Global inequality in environmental conditions underpinning human rights

This repository contains the R code and data to reproduce the findings and figures in "Global inequality in environmental conditions underpinning human rights". 

The analysis quantifies the exposure of populations to a suite of global environmental stressors, identifying the dominant Threats to Human Rights (THRs) at the global and subnational levels. A systematic literature review, aided by LLMs, was also conducted to contextualize the findings.

Repository Contents

This repository is structured as an RStudio Project (.Rproj).

Scripts (Root Directory):

01_process_raw_data.r: Ingests and standardizes raw source data.

02_match_rasters.r: Aligns all raster layers to a common global grid.

03_analysis_main.r: Performs core THR calculations and data integration.

04_generate_plots.R: Generates the final bivariate maps.

utils.r: Contains helper functions for bivariate plotting.

05_literature_review.py: A Python script to perform the LLM-aided systematic literature review, from abstract screening to downloading included papers.

Data (/data):

data/raw/: Original, unmodified source datasets (shapefiles, NetCDFs, Web of Science exports, etc.).

data/processed/: Script-generated outputs, including the /matched_rasters/ folder and the final main_analysis_df_standard.RData file.

Data Availability Note: The original, raw source datasets are not redistributed in this repository. Please refer to the accompanying manuscript for a complete list of all data sources and citations. The supplementary materials also include detailed instructions for reproducing the initial data export from the Web of Science database.

Results (/results):

results/figures/: Final figures and maps as PNG files.

Replication

Environment:

R: The analysis was performed with R (v4.x) and RStudio. All R package dependencies (e.g., terra, data.table, ggplot2) are managed within the scripts via the pacman package.

Python: The literature review script requires Python 3.x. All Python package dependencies (e.g., pandas, google-generativeai) should be installed via pip.

Workflow:

Open the .Rproj file in RStudio.

Execute the numbered R scripts (01 through 04) in the root directory sequentially.

To reproduce the literature review, execute the 05_literature_review.py script in a Python environment.

Final outputs will be generated in the /results directory.

Sensitivity Analyses: Different sensitivity analyses can be run by modifying the boolean flags (e.g., th_loose, th_restr) at the top of the 03_analysis_main.r script. Changing these flags will use different sets of input data layers for the main calculations.