Downscaled gridded global dataset for Gross Domestic Product (GDP) per capita at purchasing power parity (PPP) over 1990-2022

These codes were used to create the 'Downscaled gridded global dataset for Gross Domestic Product (GDP) per capita at purchasing power parity (PPP) over 1990-2022' that is published in Kummu et al (2024). Please do cite this paper if using the codes.

Kummu, M., Kosonen, M. & Masoumzadeh Sayyar, S. 2025. Downscaled gridded global dataset for gross domestic product (GDP) per capita PPP over 1990–2022. Scientific Data 12: 178. https://doi.org/10.1038/s41597-025-04487-x

The input data needed to run the code is available in the repository: https://doi.org/10.5281/zenodo.10976733. Please extract the zip-file (code_input_data.zip) under the same folder with the code.

The final results are within the same repository.

The code is numbered with the order it should be run. Below each code is briefly explained. We used R (version 4.3.2) to develop the code, except for downscaling which was done using Matlab (version 2023b).

0_install_packages.R: install the needed packages

1_gdp_prepare_adm0.R: puts together admin 0 level (national) data, interpolates and extrapolates the missing values

2_gdp_prep_adm1.R: puts together admin 1 level (sbunational) data, interpolates and extrapolates the missing values

3_gdp_prep_spatial.R: combines the admin 0 and admin 1 level to a global grid and gpkg file

4_downscaling_train.R: prepares the training data for downscaling model

5_downscaling_predict.R: prepares the data for downscaling prediction to admin 2 level

Downscaling_Matlab_Codes (under this folder, with own readme file): downscaling admin 1 level data to admin 2 level

6_gdpTotal.R: estimates the total GDP PPP, using admin 2 level GDP per capita (from 5_downscaling_predict.R) and population count data

7_gdp_plot_maps.R: script to plot maps shown in the manuscript

8_gdp_metadata_collect.R: puts together the metadata for the study

9_adm2_validation.R: validation of admin 2 level data

10_downscalingErrorToMap.R: creates downscaling error map

11_storeFinalFiles.R: store final files

functions: the functions used in the scripts above are in this folder

The input data needed to run the code is available in the repository: https://doi.org/10.5281/zenodo.10976733. Please extract the zip-file (code_input_data.zip) under the same folder with the code.

For more information, please contact Matti Kummu ([email protected])