Remote sensing of chlorophyll concentrations in a turbid shelf sea

Abstract

The measurement of phytoplankton (and therefore chlorophyll concentration) in
turbid waters using satellite based instrumentation is a challenging task. Traditional algorithms tend to overestimate chlorophyll concentrations due to the
unpredictable interference and enhancement of the overall water leaving signal by sediments and other non-chlorophyllous substances in the water. This study sets out to find a new satellite-based method for the retrieval or chlorophyll
concentrations in turbid waters, allowing for more accurate measurements of oceanic primary production to be made, ultimately assisting in the study of global
carbon fluxes and drawdown.
Initially, basic optical theory was employed to model the effects of sedimentary
and coloured dissolved organic matter (CDOM) upon traditional blue:green ratio
chlorophyll algorithms, such as Sea WiFS OC4v4. It was found that whilst mineral
suspended solids (MSS) often acted to enhance the water leaving signal, resulting
in overestimates of chlorophyll concentration in turbid waters, at times it resulted
in an underestimation of chlorophyll levels, an unexpected, and unpredictable
finding. Findings from the model were validated in a series of controlled experiments, where the model results were replicated.
Since the optical model and experiments confirmed the fact that algorithms such
as OC4v4 (developed for use in open ocean waters, but often used on a global
scale) would always perfonn poorly in coastal waters (with inconsistent and
unpredictable results), two new algorithms were developed to improve upon this
fact. To begin with, the effects of MSS upon the water leaving signal were quantified (using reflectance at 555nm), and accounted for within a new blue:green reflectance ratio algorithm, based upon OC4v4. An improvement in
chlorophyll retrieval accuracy was observed.
A second algorithm was then developed, this time having tuneable parameters that were able to adjust the algorithm at every single data point. Again, an
improvement in chlorophyll retrieval was noted.
Thirdly, a technique was developed that used inherent optical properties and
optical theory to estimate concentrations of chlorophyll, MSS and CDOM in
turbid waters. Whilst the technique was relatively successful, it was not able to
improve upon the accuracy achieved by the tuneable algorithm.
In conclusion it was found that although improvements could be made in terms of chlorophyll estimation in optically complex waters, the problem still remains. The unpredictable and highly variable nature of such water bodies means that a global solution is an unlikely prospect. Nevertheless, this study has resulted in the development of several new techniques to quantify chlorophyll concentration in the Irish Sea, significantly improving upon existing methods.

Date of Award	Feb 2009
Original language	English
Awarding Institution	University of Wales, Bangor
Sponsors	NERC
Supervisor	David Bowers (Supervisor) & Elizabeth Mitchelson-Jacob (Supervisor)