Benthic Reflectance Maps for Coral Sea Atolls from Sentinel-2 and Bathymetry

A method for estimating seafloor reflectance up to 50-60 meters depth in clear waters using Sentinel-2 satellite imagery and bathymetry data. The dataset includes case studies for three reef complexes in the Coral Sea and part of the Great Barrier Reef. This work was developed under the NESP MaC 2.3 project by AIMS and shared via the Australian Ocean Data Network.

Use Cases

• Mapping sand versus vegetation on the seafloor based on estimated benthic reflectance.
• Assessing the effectiveness of different Sentinel-2 spectral bands for deep-water benthic mapping.
• Testing the sensitivity of reflectance estimates to degraded input data, such as low-resolution bathymetry.
• Providing a visual reference for manual mapping of oceanic vegetation on coral atolls.

Strengths

• Method enables benthic reflectance mapping at depths of 50-60 meters, deeper than typical multispectral approaches limited to 10-20 meters.
• Includes case studies for multiple reef complexes: Flinders, Holmes, Lihou, and parts of the Great Barrier Reef.
• Analysis performed for the first four spectral bands of Sentinel-2 imagery (B1-B4) to assess relative effectiveness.

Limitations

• Method only works where water is clear, matching bathymetry data exists, and there are enough flat light/dark areas for calibration.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.

Provenance

Source

Australian Ocean Data Network (data_gov_au), associated with AIMS and the NESP MaC 2.3 project.

Collection Method

Benthic reflectance estimated pixel-by-pixel from a combination of Sentinel-2 satellite imagery and independent bathymetry digital elevation models.

Freshness

Last updated 2026-07-13 13:57:38.750505; freshness should be verified.

Geography

Coral Sea (Flinders Reefs, Holmes Reefs, Lihou Reefs) and a section of the Great Barrier Reef north-east of Townsville.