BathyUNet++ : a bathymetry retrieval framework for high-resolution low-SNR SuperDove imagery
Year: 2026
Type: article
Abstract: High-spatial-resolution (HR) bathymetry is crucial for ensuring navigation safety and effective environmental management in shallow waters. However, traditional bathymetry retrieval from medium-resolution satellites yields large uncertainties under complex substrates and strong spatial heterogeneity. HR sensors, such as PlanetScope SuperDove (~3.7 m), can resolve fine-scale features but often have low signal-to-noise ratio (SNR) and limited spectral bands, leading to noisy bathymetry retrievals. To address this challenge, we developed BathyUNet++, a UNet++ architecture enhanced with spatial and channel squeeze & excitation (scSE) attention modules, for bathymetry retrieval from SuperDove imagery. While this architecture effectively leverages spatial context and complementary spectral information, it requires dense training patches. Traditional depth measurements provide sparse spatial coverage, and Airborne LiDAR Bathymetry (ALB) remains geographically limited. Therefore, we adopted a multi-source training strategy that integrates high-resolution ALB data with extensive Landsat 8/9 derived bathymetry to improve the spatial coverage and diversity of training data for SuperDove imagery. In addition, we designed a Central-Region Mean Squared Error (CMSE) that focuses on each patch's center valid receptive field to reduce instability and checkerboard artifacts often seen in satellite products generated via patch-based processing. Independent validation shows that BathyUNet++ achieves high-accuracy bathymetry retrievals, with median absolute percentage differences below 15% compared with independent depth measurements from ICESat-2 and ALB. Notably, BathyUNet++ substantially reduces noise compared with bathymetry maps derived from pixel-based algorithms, yielding spatially consistent bathymetry maps. The proposed framework overcomes the limitations of low-SNR data and can be readily adapted to other HR sensors, offering a promising solution for the broader application of HR bathymetry retrieval in optically shallow waters.
Source: Figshare
Language: English
Subfield: Oceanography
Field: Earth and Planetary Sciences
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225809322
FWCI:
Cites:
Cited by:
Related to: