Border-city RPV Study

This repository contains the processed data and analysis scripts for the manuscript:

“Border cities reveal how rooftop solar opportunity becomes urban deployment”

The repository is provided to support peer review, reporting-summary documentation, and reproducibility of the processed-data analyses reported in the manuscript.

Overview

This study compares rooftop photovoltaic (RPV) deployment across selected neighboring border-city pairs to examine how rooftop solar opportunity is translated into urban deployment under different local economic, policy, and administrative conditions.

The analyzed city pairs are:

• San Diego, United States — Tijuana, Mexico
• El Paso, United States — Ciudad Juárez, Mexico
• Hong Kong — Shenzhen, China
• Singapore — Johor Bahru, Malaysia
• Vienna, Austria — Bratislava, Slovakia
• Monaco — Nice, France

The repository includes processed rooftop PV/building data, policy-friction scores, city-level economic inputs, and scripts for reproducing the main analytical summaries and figures.

Repository structure

Border-city-RPV-study/
│
├── data/
│   ├── Building_PVs/        # Processed rooftop PV/building tables and validation summaries
│   ├── PV_Eco_model/        # Processed economic-model inputs and outputs
│   ├── Policy_frictions/    # Policy-friction codebook and scoring tables
│   └── city_economic/       # City-level economic and deployment data
│
├── scripts/
│   ├── econimic_model.py
│   └── plot_*.py            # Figure-generation scripts
│
└── README.md

Note: the filename econimic_model.py is retained as currently used in the repository.

Data

The repository contains processed data tables used for the manuscript analyses.

data/Building_PVs/

Processed rooftop PV and building-level summary tables, including city-level image coverage summaries, rooftop PV utilization tables, and validation/benchmark summaries.

data/Policy_frictions/

Policy-friction scoring files used to summarize revenue and administrative barriers to rooftop PV deployment.

The friction indicators are:

• A: Export compensation friction
• B: Export constraint friction
• C: Settlement complexity friction
• D: Policy uncertainty friction
• E: Small-system approval friction
• F: Building/planning approval friction
• G: Grid study/fee friction
• H: Professional credential friction

Each indicator is scored from 0 to 3, where higher values indicate greater friction.

data/PV_Eco_model/

Processed inputs and outputs for the standardized rooftop PV economic model, including assumptions on installed cost, electricity price, export compensation, PV yield, self-consumption, degradation, and O&M.

data/city_economic/

City-level economic and deployment indicators used in the city-pair comparison and figure-generation scripts.

Software requirements

The analysis scripts are written in Python and require a standard scientific Python environment.

Recommended:

Python >= 3.10
pandas
numpy
matplotlib
pillow

Install dependencies with:

pip install pandas numpy matplotlib pillow

Reproducing the analysis

Clone the repository:

git clone https://github.com/PEESEgroup/Border-city-RPV-study.git
cd Border-city-RPV-study

Run the economic model:

python scripts/econimic_model.py

This generates:

scripts/economic_analysis_results.csv
scripts/economic_figures/

The economic model reports standardized city-level rooftop PV metrics, including CAPEX, LCOE, NPV, IRR, simple payback, discounted payback, compensation ratio, and year-1 cash-flow quantities.

To reproduce figures, run the relevant plotting scripts in scripts/. For example:

python scripts/plot_border_city_pv_friction_heatmap.py --help
python scripts/plot_city_rpv_utilization_within_pair_hbar.py --help
python scripts/plot_city_roofsize_pv_adoption.py --help
python scripts/plot_capex_vs_profitability_citypair_scatter.py --help

Use the --help flag to check the required input and output arguments for each script.

Reproducibility notes

The included scripts are deterministic when run with the provided processed input tables. No random sampling, stochastic model training, or GPU computation is required for the processed-data analyses included here.

For reproducibility, users are encouraged to record:

python --version
pip freeze > environment_freeze.txt
git rev-parse HEAD

Scope and limitations

This repository is a trimmed reproducibility package. It includes reusable processed data and code, but does not include all raw upstream materials.

Not included:

• raw high-resolution aerial or satellite imagery;
• trained segmentation checkpoints;
• large intermediate geospatial files;
• provider-restricted datasets that cannot be redistributed;
• raw annotation files if restricted by license or size.

The economic model is intended as a standardized comparative diagnostic across cities, not as site-specific financial advice. Policy-friction scores summarize conditions at the time of analysis and should be updated if regulations change.

Data and code availability

The processed data and analysis scripts required to reproduce the reported city-level comparisons, policy-friction summaries, economic diagnostics, and figure-generation workflows are available in this repository.

Raw imagery, restricted geospatial data, and model checkpoints are not redistributed because of licensing and file-size constraints. The manuscript and supplementary information describe the data sources, processing procedures, and validation summaries used to generate the processed analytical tables.

License

This project is covered under MIT license.

Contact

For questions about the data, code, or manuscript, please contact the corresponding author listed in the manuscript.