Ngebe, SNaidoo, LVan Deventer, HeidiTsele, PQabaqaba, M2026-04-232026-04-232026-042673-6187https://doi.org/10.3389/frsen.2026.1812294http://hdl.handle.net/10204/14796Quantifying carbon stocks from the above-ground biomass (AGB) of wetland vegetation across seasons is crucial for assessing ecosystem resilience to anthropogenic and climate pressures. This study aimed to assess differences between summer and winter in aboveground carbon (AGC) of palustrine wetland vegetation using Sentinel-1 and Sentinel-2 data. The Random Forest (RF) and Support Vector Regression (SVR) were implemented with variable importance selection to develop an optimal model from remote sensing derived modelling scenario combinations. Modelling scenarios included field measured Leaf Area index and different combinations of (i) Sentinel-2 derived variables namely vegetation indices (VIs) and reflectance bands, and (ii) Sentinel-1 grey-level co-occurrence matrices, backscatter band ratios, and backscatter channels. Results indicated significant seasonal variation (p < 0.05) with higher total teal carbon in summer (155.1 g C/m2) than winter (115.8 g C/m2). Large macrophytes particularly Phragmites australis stored the highest carbon (93.04 g C/m2 in summer; 78.37 g C/m2 in winter). Sentinel-1-derived models outperformed Sentinel-2-based models for both seasons, achieving R2 of 0.7–0.8, RMSE of 39.9–69.6 g·m-2, and relative RMSE of 17.3%–21.3%. RF consistently performed better than SVR. Thus, seasonal monitoring of teal carbon provide valuable insights of wetlands vegetation contribution in carbon accounting and sequestration.FulltextenAbove-ground biomassCarbon sequestrationEssential biodiversity variablesSentinel-1Sentinel-2Articlen/a