This is a Preprint and has not been peer reviewed. This is version 1 of this Preprint.
Directional Centroid Trajectories Reveal Shifting Fire Activity Across Brazilian Biomes
Downloads
Authors
Abstract
We utilized directional centroid trajectories to examine how Brazil’s fire regimes in natural vegetation and anthropogenic land use have spatially reorganized over the last four decades. Annual area-weighted centroids were derived separately for natural and anthropogenic burned patches from the MapBiomas Fire Collection 4 (1985–2024), and their interannual displacements were quantified using great-circle distance and azimuth. Centroids associated with natural vegetation and anthropogenic use formed persistently distinct spatial clusters, with a mean separation of 313 km and low kernel overlap (KOI = 0.17), indicating structurally different spatial domains of fire activity. Despite this separation, most interannual centroid displacements were directionally ‘aligned’ or ‘moderately aligned’, with truly ‘divergent’ years being rare and directionally scattered. Major climate anomalies (El Niño, La Niña, ZCAS-related droughts) showed no consistent association with directional divergence. Biome-level burned-area deltas revealed coherent but noisy patterns by ‘Attraction Region’, suggesting that centroid trajectories integrate multi-biome redistributions of fire without a single dominant driver. An exceptional convergence of natural and anthropogenic centroids in 2024 was linked to a record increase in burned natural vegetation in the Amazon, illustrating how extreme drought can temporarily homogenize otherwise distinct fire regimes.
DOI
https://doi.org/10.31223/X5W47H
Subjects
Life Sciences
Keywords
Wildfire dynamics, Climate extremes, land use change, Spatial fire patterns, MapBiomas, Climate Extremes, land use change, Spatial fire patterns, MapBiomas
Dates
Published: 2025-12-10 02:10
Last Updated: 2025-12-10 02:10
License
CC BY Attribution 4.0 International
Additional Metadata
Conflict of interest statement:
None
Data Availability (Reason not available):
Data and code associated with this preprint are publicly available on the Open Science Framework (OSF; DOI: 10.17605/OSF.IO/WA326).
There are no comments or no comments have been made public for this article.