Coal plants persist as a large barrier to the global solar energy transition - Dataset

Interactive Dashboard

Explore the facility-level PV generation and aerosol-loss dataset interactively at pvfacilitymap.uk.

Click the preview to open the live dashboard in a new tab. The dashboard complements this repository: this repo provides the raw data and reproducible scripts, while the dashboard offers a visual, exploratory entry point.

Overview

This repository provides example code and reproducible plotting workflows for the global facility-level solar PV dataset associated with the manuscript:

Coal plants persist as a large barrier to the global solar energy transition.

The dataset used here includes:

• a global PV facility inventory (PV_ID, latitude, longitude, country, year, area)
• yearly facility-level PV generation and aerosol-related loss files (PV_facility_generation_year_YYYY.csv)

The goal of this repo is practical reproducibility:

• understand dataset structure quickly
• reproduce selected manuscript-style aggregate figures
• provide a clean codebase that can be extended in future

Data Source and Citation

This repository contains convenient working copies of selected files for examples. For scientific use and citation, the authoritative dataset source is Zenodo:

• Zenodo record: https://zenodo.org/records/18794231
• Zenodo DOI: 10.5281/zenodo.18794231

Please cite:

1. The Zenodo dataset DOI (10.5281/zenodo.18794231).
2. The manuscript.
3. This repository version/commit (if using scripts from this repo).

Code license: MIT (see LICENSE). Dataset license and usage terms: follow the Zenodo record.

Citation metadata is also provided in CITATION.cff.

Repository Layout

• data/: dataset files used by example scripts
• scripts/examples/: runnable analysis and figure-replication scripts
• outputs/tables/: generated summary tables
• outputs/figures/: generated figures
• docs/: data dictionary and release notes
• metadata/: metadata and checksum files
• src/: reusable modules (for future expansion)
• tests/: tests

Quick Start

python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

Run core scripts:

python scripts/examples/interpret_global_pv_parquet.py --save-output
python scripts/examples/plot_aerosol_loss_vs_new_generation.py
python scripts/examples/plot_country_aerosol_loss_2023.py

Figure Replication

• Global aerosol-loss vs new PV generation:

python scripts/examples/plot_aerosol_loss_vs_new_generation.py --year-start 2017 --year-end 2023

• Manuscript Fig.2f style (2023 country comparison):

python scripts/examples/plot_country_aerosol_loss_2023.py

Expected Sanity Checks

If the input files are unchanged, key outputs should be close to:

• plot_country_aerosol_loss_2023.py
  • China aerosol-loss share: ~54.91%
  • India aerosol-loss share: ~12.99%
  • USA aerosol-loss share: ~9.63%

Output File Policy

• Figure files in outputs/figures (.png, .pdf) are versioned for reference.
• Table outputs in outputs/tables remain ignored by Git (except .gitkeep).
• Regenerate outputs by rerunning scripts locally.

Included Data Files

• data/global_pv_facility_inventory.csv
• data/pv_generation_losses/PV_facility_generation_year_2017.csv
• data/pv_generation_losses/PV_facility_generation_year_2018.csv
• data/pv_generation_losses/PV_facility_generation_year_2019.csv
• data/pv_generation_losses/PV_facility_generation_year_2020.csv
• data/pv_generation_losses/PV_facility_generation_year_2021.csv
• data/pv_generation_losses/PV_facility_generation_year_2022.csv
• data/pv_generation_losses/PV_facility_generation_year_2023.csv

Release Notes

See CHANGELOG.md.