Extensive ecosystem degradation along the coastal areas of Ghana highlights the need for targeted landscape restoration. This study identified areas with high restoration potential by evaluating environmental, climatic, and anthropogenic determinants of mangrove distribution. We tested multiple habitat suitability models, including Random Forest (RF), Generalized Additive Model (GAM), Generalized Linear Model (GLM), Maximum Entropy (MaxEnt), and an ensemble approach. Mangrove occurrence records were compiled from field observations, drone-derived data, and publicly available biodiversity and mapped datasets. The ensemble GAM-RF model emerged as the most robust model, with the highest predictive performance. Hydrological and topographic factors were the strongest predictors of mangrove habitat suitability, particularly proximity to rivers and the coastline, low-elevation terrain (<5 m asl), and temperature-related bioclimatic variables, while precipitation metrics and anthropogenic proxies played secondary roles. Using the ensemble approach, the estimated potentially suitable habitat ranged from 417 to 1313 km², indicating that substantial areas remain available for mangrove restoration. These findings align with ongoing national and international restoration initiatives, including the coastal restoration plan, and have important implications for coastal protection, carbon sequestration, fisheries habitat, and local livelihoods. Our findings demonstrate that multi-model habitat suitability analyses can guide spatially targeted, and evidence-based restoration planning.

Remote sensing is essential for mapping mangrove forest ecosystems, yet accuracy and spatial resolution remain limited, particularly in Africa, where mangroves are fragmented, reference data and cloud-free images are scarce. The state-of-the-art of mangrove forest mapping relies on native-resolution imagery using machine learning. This study evaluates deep learning methods using aggregated multi-date satellite imagery to improve mangrove mapping and national-scale area estimation. Super-resolution and semantic segmentation techniques were applied to Sentinel-2 (S2) and Sentinel-1 (S1) images, with training data derived from field observations and high-resolution drone and satellite imagery. Coastal regions of Tanzania and Ghana were used as representative study areas for eastern and western Africa, respectively. Results show strong classification performance, with S2 imagery alone achieving an Intersection over Union (IoU) of 85.8%. Adding S1 provided modest gains (IoU 86.3%), while super-resolution significantly improved the accuracy (IoU 88.5%). The combined super-resolved S2 and S1 dataset yielded the highest performance (IoU 89.2%). Wall-to-wall mapping and statistically rigorous validation and area estimation produced accurate results with quantified uncertainty. Tanzania showed high consistency (F1 = 96%; 1207 km², 95% CI [1151.7, 1263]), while Ghana exhibited slightly higher uncertainty (F1 = 92%; 92.6 km², 95% CI [87.5, 97.7]). Integrating deep learning with multi-date, multi-sensor fusion and super-resolution enables accurate mangrove mapping and area estimation for improved coastal management and restoration.

Remote sensing is essential for mapping mangrove forest ecosystems, yet accuracy and spatial resolution remain constrained, particularly in Africa, where mangroves are fragmented, reference data and cloud-free images are scarce. The state-of-the-art of mangrove forest mapping relies on native-resolution, single-date imagery and machine learning. This study evaluates deep learning methods using aggregated multi-date satellite imagery to improve mangrove mapping and national-scale area estimation. Semantic segmentation and super-resolution techniques were applied to Sentinel‑2 (S2) and Sentinel‑1 (S1) images, with training data derived from field observations and high-resolution drone and satellite imagery. Coastal regions of Tanzania and Ghana were used as representative study areas for eastern and western Africa, respectively. Results show strong classification performance, with S2 imagery alone achieving an Intersection over Union (IoU) of 85.8%. Adding S1 provided modest gains (86.3%), while super-resolution significantly improved the accuracy (88.5%). The combined super-resolved S2 and S1 dataset yielded the highest performance (89.2%). Wall-to-wall mapping and statistically rigorous validation and area estimation produced accurate results with quantified uncertainty. Tanzania showed high consistency (F1 = 96%; 1207 km², 95% CI [1151.7, 1263]), while Ghana exhibited slightly higher uncertainty (F1 = 92%; 92.6 km², 95% CI [87.5, 97.7]). Integrating deep learning with multi-date, multi-sensor fusion and super-resolution enables accurate mangrove mapping and area estimation for improved coastal management and restoration.