Scholarlink

<p>Accurate and up-to-date bathymetric data are essential for the management and spatial planning of transitional coastal environments, yet conventional survey methods remain costly and difficult to implement at large spatial scales and high temporal frequency. This study investigates the potential of Satellite-Derived Bathymetry (<i>SDB</i>) as a cost-effective monitoring tool, using the Venice Lagoon as a morphologically and optically complex test site. Two methodological axes were developed. The first applies a stratified log-ratio Band Ratio Method (Stumpf et al. 2003) using Sentinel-2 multitemporal imagery, calibrated against the 2002 full-lagoon bathymetric dataset (Sarretta et al. 2010), achieving a Root Mean Square Error (RMSE) of 1.09&#xa0;m and an R² of 0.87 across the entire depth range. The second axis refines this approach through advanced radiometric pre-processing, including sun-glint correction and clean-water compositing, combined with a Random Forest substrate classification and a stratified <i>SDB</i> calibration for sandy versus muddy seafloors. The resulting <i>SDB</i> was further integrated with Kriging with External Drift (<i>KED</i>), using the 2013 CNR-ISMAR (National Research Council Institute of Marine Sciences) multibeam survey as ground truth. This hybrid <i>SDB-KED</i> model achieved an R² of 0.73, an RMSE of 2.55&#xa0;m, and a near-zero bias (-0.09&#xa0;m) over more than 320,000 validation points, demonstrating effective bathymetric reconstruction across the full lagoon depth range. Residual uncertainties are concentrated at channel margins and in areas exceeding 15&#xa0;m depth, where optical signal extinction limits satellite penetration. The proposed framework, implemented entirely on open-source data and cloud computing platforms, offers a replicable and scalable protocol for supporting spatial planning in dynamic coastal and marine environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Satellite-derived bathymetry in optically complex waters: testing an SDB–KED hybrid approach in the Venice Lagoon

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月20日

查看全文
  • Federica Gerla,
  • Caterina Balletti

摘要

Accurate and up-to-date bathymetric data are essential for the management and spatial planning of transitional coastal environments, yet conventional survey methods remain costly and difficult to implement at large spatial scales and high temporal frequency. This study investigates the potential of Satellite-Derived Bathymetry (SDB) as a cost-effective monitoring tool, using the Venice Lagoon as a morphologically and optically complex test site. Two methodological axes were developed. The first applies a stratified log-ratio Band Ratio Method (Stumpf et al. 2003) using Sentinel-2 multitemporal imagery, calibrated against the 2002 full-lagoon bathymetric dataset (Sarretta et al. 2010), achieving a Root Mean Square Error (RMSE) of 1.09 m and an R² of 0.87 across the entire depth range. The second axis refines this approach through advanced radiometric pre-processing, including sun-glint correction and clean-water compositing, combined with a Random Forest substrate classification and a stratified SDB calibration for sandy versus muddy seafloors. The resulting SDB was further integrated with Kriging with External Drift (KED), using the 2013 CNR-ISMAR (National Research Council Institute of Marine Sciences) multibeam survey as ground truth. This hybrid SDB-KED model achieved an R² of 0.73, an RMSE of 2.55 m, and a near-zero bias (-0.09 m) over more than 320,000 validation points, demonstrating effective bathymetric reconstruction across the full lagoon depth range. Residual uncertainties are concentrated at channel margins and in areas exceeding 15 m depth, where optical signal extinction limits satellite penetration. The proposed framework, implemented entirely on open-source data and cloud computing platforms, offers a replicable and scalable protocol for supporting spatial planning in dynamic coastal and marine environments.