OlalekanAlagbe/nigeria-flare-analysis
Domain:
environment and energygeospatialRecord type:
project
Creator:
Ola
Host:
# Nigeria Gas Flare Detection & Emissions Analysis (2012–2024)
A data science portfolio project analysing 13 years of World Bank VIIRS satellite data to map, quantify, and model Nigeria's gas flaring crisis — with particular focus on Seplat Energy's Flares Out programme and the strategic implications of the December 2024 SEPNU acquisition.
**Live site:** olalekanalagbe.github.io
---
## Screenshots
> After deploying to GitHub Pages, replace this note with a screenshot of the live page.
---
## Setup
### 1. Install dependencies
```bash
pip install pandas openpyxl scikit-learn matplotlib seaborn
# Optional: for basemap tile layer on Figure 7
pip install geopandas contextily
```
### 2. Generate figures and data
Place `NIGERIA FLARING DATA.xlsx` in the project root, then run:
```bash
python analysis.py
```
This creates:
- `figures/` — 8 PNG files (figure1_trend.png … figure8_ml.png)
- `chart_data.json` — summary statistics and aggregated series
### 3. View locally
Open `index.html` directly in any modern browser. No build step or local server required.
### 4. Deploy to GitHub Pages
```bash
git init
git add index.html figures/ chart_data.json analysis.py README.md
git commit -m "Initial publish: Nigeria gas flare analysis"
git remote add origin github.com
git push -u origin main
```
Then in the GitHub repository settings, enable **Pages → Deploy from branch → main / (root)**.
---
## Tech Stack
| Component | Tool |
|-----------|------|
| Data processing | Python 3, pandas, numpy |
| Machine learning | scikit-learn RandomForestRegressor |
| Visualisation | matplotlib, seaborn |
| Geospatial | geopandas, contextily (optional) |
| Web page | Vanilla HTML/CSS, Google Fonts |
| No JavaScript charts | All figures are static matplotlib PNGs |
---
## Key Results
| Metric | Value |
|--------|-------|
| Total observations | 2,264 |
| Years covered | 2012–2024 (13 years) |
| Uniqu …