A Simple Cloud-Native Spectral Transformation Method to Disentangle Optically Shallow and Deep Waters in Sentinel-2 Images

Confronting turbidity, the major challenge for satellite-derived coastal bathymetry

Monitoring the complex seafloor morphology that drives the functioning of shallow coastal ecosystems is vital for assessing marine activities. Satellite-derived bathymetry (SDB) can provide a crucial dataset for creating the bathymetry maps needed to understand hazards and impacts produced by climate change in vulnerable coastal zones. SDB is effective in clear water, but still has limitations in application to areas with some turbidity. Here, using the twin satellites Sentinel-2A/B, we integrate water quality information from the satellite with a multi-temporal compositing method to demonstrate a potential for comprehensively operational bathymetric mapping over a range of environments. The automated compositing method diminishes the turbidity impact in addition to inferring the maximum detectable depth and removing optically deep-water areas. Examining a wide range of conditions along the Caribbean and eastern coast of the U.S. shows detailed bathymetry as deep as 30 m at 10 m spatial resolution with median errors <1 m when compared to high-resolution lidar surveys. These results demonstrate that the model adopted can provide useful bathymetry in areas that do not have consistently clear water and can be extended across multiple geographic regions and optical conditions at local, regional, and national scales.

Drone-Based Bathymetry Modeling for Mountainous Shallow Rivers in Taiwan Using Machine Learning

The river cross-section elevation data are an essential parameter for river engineering. However, due to the difficulty of mountainous river cross-section surveys, the existing bathymetry investigation techniques cannot be easily applied in a narrow and shallow field. Therefore, this study aimed to establish a model suitable for mountainous river areas utilizing an unmanned aerial vehicle (UAV) equipped with a multispectral camera and machine learning-based gene-expression programming (GEP) algorithm. The obtained images were combined with a total of 171 water depth measurements (0.01–1.53 m) for bathymetry modeling. The results show that the coefficient of determination (R2) of GEP is 0.801, the mean absolute error (MAE) is 0.154 m, and root mean square error (RMSE) is 0.195 m. The model performance of GEP model has increased by 16.3% in MAE, compared to conventional simple linear regression (REG) algorithm, and also has a lower bathymetry retrieval error both in shallow (<0.4 m) and deep waters (>0.8 m). The GEP bathymetry retrieval model has a considerable degree of accuracy and could be applied to shallow rivers or near-shore areas under similar conditions of this study.