Mapping global sandy beach slope using multi-source remote sensing data and ensemble learning

Sandy beach slope is an essential parameter characterizing coastal geomorphology and nearshore hydrodynamic processes, which is of great significance for coastal erosion monitoring, flood risk assessment, and coastal vulnerability analysis. However, global-scale sandy beach slope mapping still faces significant challenges due to the high costs and limited coverage of in-situ measurements, as well as the insufficient spatial resolution of existing global topographic products. Therefore, this study proposes a global sandy beach slope mapping method based on multi-source remote sensing data and ensemble learning. This research integrates Sentinel-1/2 remote sensing imagery, sandy beach morphology data, marine environmental data, climate reanalysis data, and human activity data to construct a multi-source feature system, and establishes a global sandy beach slope estimation framework using a Stacking ensemble learning model. The results indicate that: (1) The Stacking model achieves the best predictive performance, with , RMSE, MAE, and Bias on the validation set being 0.8879, 0.0250, 0.0167, and -0.0003, respectively, comprehensively outperforming other models across all metrics; (2) Feature importance analysis reveals that beach width (Width) and area (Area) are the dominant factors influencing sandy beach slope, while the contribution proportion of remote sensing features in the ensemble model ranges from 13.7% to 35.5%; (3) Global sandy beaches are predominantly characterized by gentle slopes (0.02–0.09), accounting for 54.15%, whereas steep beaches (0.18–0.27) and exceptionally cliffed beaches (>0.27) only account for 4.97% and 2.12%, respectively, and are primarily distributed along the Mediterranean coast of Europe and the west coast of South America. This study demonstrates the effectiveness of integrating multi-source remote sensing data and ensemble learning methods in global coastal topographic mapping, providing vital data support for global coastal geomorphological research, coastal erosion monitoring, and coastal risk assessment.