This product is a raster map in GeoTIFF format with a spatial resolution of 10 m/px, representing the estimated Aboveground Biomass (AGB) of mangroves in the northwestern sector of the Vía Parque Isla de Salamanca (VIPIS) in 2025. The area is located on the Caribbean coast of Colombia, in the department of Magdalena. The dataset was generated through the integration of Sentinel-1 (radar) and Sentinel-2 (optical) satellite imagery, canopy height data, and field measurements from 20 plots, complemented with additional measurements from Colombian mangroves with similar ecological characteristics available in SIGMA.The methodology employed machine learning techniques (Random Forest) to train and validate predictive AGB models, using validation metrics such as the Root Mean Square Error (RMSE = 152.545). Satellite imagery was processed in Google Earth Engine to generate dry-season mosaics (January–April 2025) and extract relevant spectral, textural, and structural variables. The mapped area covers approximately 6,000 hectares of mangroves within the northwestern boundaries of VIPIS, projected in UTM Zone 18N (EPSG:32618). This product is part of the Quantifying Colombian Mangroves Aboveground Biomass and Carbon Content – VIPIS Case (2025) project, developed by CTTC and INVEMAR.

This product is a raster map in GeoTIFF format with a spatial resolution of 10 m/px, representing the estimated Aboveground Biomass (AGB) of mangroves in the northwestern sector of the Vía Parque Isla de Salamanca (VIPIS) in 2025. The area is located on the Caribbean coast of Colombia, in the department of Magdalena. The dataset was generated through the integration of Sentinel-1 (radar) and Sentinel-2 (optical) satellite imagery, canopy height data, and field measurements from 20 plots, complemented with additional measurements from Colombian mangroves with similar ecological characteristics available in SIGMA.The methodology employed machine learning techniques (Random Forest) to train and validate predictive AGB models, using validation metrics such as the Root Mean Square Error (RMSE = 152.545). Satellite imagery was processed in Google Earth Engine to generate dry-season mosaics (January–April 2025) and extract relevant spectral, textural, and structural variables. The mapped area covers approximately 6,000 hectares of mangroves within the northwestern boundaries of VIPIS, projected in UTM Zone 18N (EPSG:32618). This product is part of the Quantifying Colombian Mangroves Aboveground Biomass and Carbon Content – VIPIS Case (2025) project, developed by CTTC and INVEMAR.