This project uses Google Earth Engine and geemap in Google Colab to analyze wildfire impact and vegetation health in Tizi Ouzou, Algeria during August 2021. It calculates: NBR (Normalized Burn Ratio) → highlights burned areas and burn severity. NDVI (Normalized Difference Vegetation Index) → assesses vegetation health.
# Wildfire Analysis: Tizi Ouzou (August 2021)
This project uses Google Earth Engine and geemap in Google Colab to analyze wildfire impact and vegetation health in Tizi Ouzou, Algeria during August 2021.
# Project Overview
This project analyzes and classifies burned vs. unburned areas caused by the August 2021 wildfires in Tizi Ouzou, Algeria using satellite imagery and machine learning.
Using Sentinel-2 satellite data from the European Space Agency and Google Earth Engine, we compute multiple spectral indices and train a Random Forest classifier to automatically detect burned areas.
The result is an interactive map showing Burned vs Unburned Areas
# 🌍 Study Area
## 📍 Tizi Ouzou, Algeria
## Wildfire period: August 20 – September 10, 2021
# 🛰️ Data Source
Satellite Imagery: European Space Agency Sentinel-2
Dataset: Google Earth Engine – COPERNICUS/S2_SR_HARMONIZED
# Spectral Indices Computed
To improve burn detection accuracy, multiple indices were calculated:
• NBR – Normalized Burn Ratio (burn severity detection)
• NDVI – Vegetation health
• NDWI – Water/moisture detection
• MSAVI – Soil-adjusted vegetation index
• SWIR/NIR Ratio – Fire damage sensitivity
These indices enhance separability between burned and unburned surfaces.
# Machine Learning Model
Algorithm: Random Forest
100 trees