REVIMAR-MAP: Benthic Marine Habitats Mapping of Brazil
Year: 2026
Type: dataset
Abstract: The REVIMAR-MAP repository provides open-source methodological standards and continuous (wall-to-wall) predictive spatial models for benthic marine habitat classification across the Brazilian continental margin. Designed to support Marine Spatial Planning (MSP) and environmental impact assessments under the guidelines of the Ministry of the Environment and Climate Change (MMA) and funded by FUNBIO, the framework employs a multiscale, top-down approach structured across three hierarchical tiers: Level 1 (L1) – Geomorphological Provinces: Macro-scale seafloor compartmentalization derived from GEBCO bathymetry, smoothed via 7×7 focal mean filtering to attenuate high-frequency artifacts. The model combines slope thresholds—distinguishing the Continental Rise (0.1° to 1.2°) from Abyssal Plains (< 0.1°)—with the Bathymetric Position Index (BPI) to delineate major physiographic features such as submarine canyons, valleys, and seamounts. Level 2 (L2) – Biological Zones: Modeling of sunlight penetration reaching the seabed using multi-year climatological median Euphotic Depth ($Z_{eu}$) from the Copernicus Marine Service (CMEMS). This layer differentiates shallow shelf environments and seamount summits into Euphotic and Mesophotic zones. In deep-water aphotic environments, it establishes bathymetric boundaries separating Bathyal (> -4000 m) from Abyssal ($\le -4000\text{ m}$) domains, alongside deep seamount isolation (< -200 m). Level 3 (L3) – Substrate & Consolidated Benthic Habitats: Spatial intersection between photic zones and a modified EUNIS Folk sediment classification scheme (Long, 2006). Ternary fractions of gravel, sand, and mud are normalized into four distinct textural classes (Coarse sediment, Mixed sediment, Sand, and Mud to muddy sand), further integrated with mapped biogenic substrates (such as rhodolith beds and coral frameworks). Substrate subdivisions are restricted to continental shelves and shallow seamount summits, retaining the physiographic identity of bathyal and abyssal features. The computational pipeline is fully automated in R using the terra and sf libraries. Raster outputs are optimized as lightweight unsigned 8-bit integer GeoTIFFs (INT1U) with embedded Raster Attribute Tables (RAT) and custom color look-up tables (CLUT) generated via linear RGB interpolation. The architecture features automated archiving with versioned DOIs via Zenodo, Quarto-based technical documentation deployed to GitHub Pages, and compliance with the Brazilian National Spatial Data Infrastructure (INDE) and the Geological Mapping of Brazil (MGB) standards.
Language: Portuguese
Subfield: Oceanography
Field: Earth and Planetary Sciences
Domain: Physical Sciences
SDG: Life on land
Open Access status: green
OpenAlex ID: W7213582002
