Project Inferno is a data processing and visualization pipeline that uses NASA’s publicly available satellite remote sensing datasets to analyze wildfire severity 🔥

Burn Watch is an experimental web app for real-time fire hotspot mapping using NASA’s FIRMS API. Built with React, Leaflet, and Turf.js, it filters and displays MODIS and VIIRS detections within a GeoJSON-defined area. Deployed on Netlify.

Data science applied in the war

Türkiye'deki yangın verilerini NASA FIRMS API'si üzerinden düzenli olarak çekmekte ve kaydetmektedir.

end-to-end pipeline to predict next-day wildfire risk from NASA FIRMS (active fires) and Meteostat weather, train LightGBM, and visualize alerts in Streamlit.

Fire Information for Resource Management System - Tool

GreenPulse India is a next-generation, real-time environmental intelligence platform designed to serve as the "digital nervous system" for India's ecological health. It bridges the gap between raw environmental sensors and actionable human intelligence by leveraging high-performance streaming technologies and Large Language Models.

Geospatial wildfire perimeter modeling system using NASA FIRMS, DBSCAN clustering, outlier filtering, and concave hull perimeter construction.

This is a world wide fire events data visualization project based on Streamlit by Python, data provided by NASA FIRMS.

A tool that facilitates access to information about hotspots in the Amazon. It generates reports and maps of hotspots in real time and for the past 7 days. The plugin does this by accessing data from NASA and INPE.

Result of a Research Initiation grant given by the São Paulo State Resarch Foundation (FAPESP) from 2025-2026. Uses Random Forest to train a wildfire prediction model by feeding environmental data (vegetation indices, temperature, land use) to a binary predictor with probabilities of occurences.

Automated retrieval of farm fire locations to facilitate stakeholders with corrective on-ground actions.

Result of a Research Initiation grant given by the São Paulo State Resarch Foundation (FAPESP) from 2025-2026. Analyzes wildfire hotspots over the Cerrado biome in São Paulo state using Kernel Density Estimates.

Quick data visualisation of thermal data for airborne campaign planning

Bristol Scientific Computing (BriSC) demonstration of using Python tools in a scientific context. This demo shows how we can use code to examine and plot global fire data.

Agentic wildfire detection and risk assessment with NASA FIRMS satellite thermal anomalies (VIIRS/MODIS), NOAA fire weather indices, EPA AirNow air quality data, and PurpleAir real-time PM2.5 sensors