Data Repository for Ignition Analysis
Description
This dataset and its accompanying documentation provide a framework for analyzing wildfire ignition points and their resulting burned areas within the Pantanal biome, specifically in the state of Mato Grosso do Sul, Brazil. The data covers the 2024 fire moratorium (starting in May 2024), a period during which a "State of Environmental Emergency" was declared due to climate conditions highly conducive to the spread of wildfires. The methodology utilizes a Geographic Information System (GIS) environment for near-real-time detection. By leveraging multi-scale remote sensing data, the study delineates precise ignition polygons and the total extent of burned areas. Key Technical Terms Ignition polygons: The exact locations where the fires originated. Burned areas: The fire "scars" left behind on the landscape. Geographic Information System (GIS): A framework for gathering, managing, and analyzing spatial data. Near-real-time: Data processing and relay that occurs almost immediately after an event. Remote sensing: The process of monitoring physical characteristics of an area from a distance (typically via satellite or aircraft).
Files
Steps to reproduce
Overview and Environmental Context This documentation outlines the methodology for monitoring forest fires within the Pantanal biome, specifically in Mato Grosso do Sul, Brazil . The data focuses on the 2024 fire prohibition period, which saw the declaration of a "State of Environmental Emergency" due to extreme climate conditions conducive to rapid fire spread. The study employs a Geographic Information System (GIS) framework to detect ignition points and delineate total burned areas in near-real-time. • Spatial Reference System: All data uses the SIRGAS 2000 coordinate system (EPSG: 4674). Methodological Workflow The process involves defining the study area, acquiring real-time hotspot data, and utilizing multi-scale satellite imagery to identify ignition polygons (the smallest initial burn areas) and subsequent burn scars. QGIS and Google Earth Engine (GEE) Integration Data acquisition is centralized in QGIS through specific cloud-based plugins: • GEE Plugins: The "Google Earth Engine" and "Data Catalog" plugins allow users to access Landsat and Sentinel-2 raster data directly within the GIS environment. • Authentication: Users must authenticate via a registered Google account and link a valid Google Cloud Project ID to manage resource allocation and billing. Remote Sensing Data Sources The methodology utilizes a tiered approach to satellite monitoring: • Active Fire Detection (Hotspots): Data is sourced from NASA’s FIRMS and the Pantanal em Alerta platform. Specifically, NOAA-20 (VIIRS) data at 375 m resolution is refined using a 187.5 m buffer analysis to pinpoint ignition sites. • High-Resolution Imagery: PlanetScope data (~3 m resolution, daily revisit) is used to detect the earliest visual evidence of fire. • Medium-Resolution Imagery: Sentinel-2 (20 m) and Landsat-8/9 (30 m) are primary tools for mapping burn scars using Short-Wave Infrared (SWIR) and Near-Infrared (NIR) bands. • Regional Satellites: Data from Amazonia-1, CBERS-4, and CBERS-4A are accessed through the Brazilian National Institute for Space Research (INPE). • Meteorology: Wind speed and direction from CPTEC/INPE, alongside GOES-R geostationary imagery, are integrated to analyze and predict the direction of fire spread. Spectral Band Compositions (RGB) To effectively distinguish burned areas from soil and healthy vegetation, the following RGB combinations are applied: Satellite Sensor RGB Composition Resolution Sentinel-2 MSI R11/G8/B4 or R12/G11/B2 20 m Landsat-8/9 OLI R6/G5/B4 or R7/G6/B2 30 m CBERS-4A WPM R2/G4/B3 8 m CBERS-4 WFI R14/G16/B15 55 m Amazonia-1 WFI R2/G4/B3 65 m
Institutions
- Universidade Catolica Dom Bosco
- Ministerio Publico