•  
  •  
 

The Plymouth Student Scientist

Document Type

Biological and Marine Sciences Article

Abstract

This study investigates the potential of reflectance band ratios, modelled on Sentinel-2’s MultiSpectral Instrument (MSI) bands 10m resolution to estimate the downwelling attenuation coefficient for planar irradiance, Kd(λ), at 490nm (Kd(490)) in Plymouth Sound. Kd(λ) is vital for characterising subsurface light climate, strongly influencing physical and biogeochemical processes in coastal and oceanic waters. In temperate shelf seas, human activities have led to loss in vitality and spatial extent of structurally complex seabed, emphasizing the need for enhanced monitoring. Satellite-based remote sensing (RS) offers spatial and temporal advantages to conventional in situ methods. There are, however two main inadequacies that persist with the satellite based approach to obtain Kd estimates in coastal and transitional waters. 1, Sensor resolution, which fails to capture the short spatial gradients exhibited, especially in narrow water bodies, and 2, the typical band ratio models employed are confounded by the presence of optically active constituents (OAC) that do not covary with phytoplankton pigments as they often do in the open ocean. Sentinel-2’s MSI represents an opportunity to overcome challenges in spatial resolution of conventional ocean colour satellites, if suitable local and regional Kd models are derived. Leveraging an extensive dataset (N=92) of coincident in situ measurements of above surface radiance, and below surface profiles of downwelling irradiance, Kd(490) estimates from 145 band ratio variations were assessed. Four promising band ratios were identified, particularly those incorporating the use of red and near-infrared (NIR) reference bands. The best-performing algorithm (R2: 0.760) is proposed for estimating Kd(490) from Sentinel-2 data at 10m spatial resolution over turbid coastal waters comparable to Plymouth Sound. If successful application and validation of the model with Sentinel-2 data is achieved, this model could be used as a tool that, for example, could support seagrass restoration site selection or waterbody eutrophication susceptibility assessments.

Publication Date

2024-07-29

Publication Title

The Plymouth Student Scientist

Volume

17

Issue

1

First Page

1

Last Page

27

ISSN

1754-2383

Deposit Date

2024-07-29

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Share

COinS