Aboveground biomass field data with Sentinel-2 texture metrics for tropical forests in Mexico

This dataset contains field-based estimates of aboveground biomass (AGB) in tropical forests of Mexico, coupled with texture metrics derived from Sentinel-2 satellite imagery. The dataset was compiled from multiple sources and contributors to support the development and evaluation of remote sensing models for biomass estimation across heterogeneous tropical ecosystems.

Field measurements of biomass originate from a combination of datasets produced by different research groups and institutions, including the National Forest and Soil Inventory (INFyS) and independent field campaigns conducted by multiple authors. The contributing authors associated with the dataset are identified through their initials (e.g., YG, JVS, DLJ-R, AGV, MANG, CP-C, JAG-C, DC, DP, GW, JAM, GI-M, OC-A, JOL-M, FM, RCN-M, RMC, JLH-S), corresponding to the researchers listed in the associated publication. These datasets were harmonized into a unified structure to ensure consistency in variables and units.

Aboveground biomass values were calculated using allometric equations following Chave et al. (2014), Ramírez et al. (2017), Martínez-Yrizar et al. (1992), Hernández-Stefanoni et al. (2014), and guidelines from CONAFOR (2021). Inclusion criteria varied among data sources, with minimum diameter at breast height (DBH) thresholds of ≥ 10 cm, ≥ 7.5 cm, ≥ 5 cm, or > 1 cm. Each record corresponds to a georeferenced plot and includes associated  vegetation type, year of measurement, plot size, inclusion criteria, biomass equation source, conservation status, and principal investigator.

For each plot location, a suite of texture metrics was extracted from Sentinel-2 imagery using gray-level co-occurrence matrices (GLCM). These metrics were computed for both near-infrared (NIR) and red (R) spectral bands and include mean, variance, homogeneity, contrast, dissimilarity, entropy, second moment, and correlation.