@phdthesis{01GX9E0QQP26V3FQGX8JSFG0KK,
  abstract     = {{Hydrologic optics and aquatic remote sensing are fields of applied optics that aim to retrieve information of state and process of aquatic systems from optical signals. Such an inference link depends on a rich set of paired optical and biogeochemical data, robust development and evaluation of algorithms and deep understanding of the underlying processes in order to support meaningful ecological interpretation. In this research we explore those different aspects by evaluating and proposing improvements for methods of field spectroscopy, describing a dataset with detailed compositional information covering a diverse range of environmental conditions in Belgian waters, and by providing algorithms to retrieve information of the phytoplankton assemblage for use with in situ optical instrumentation and remote sensors.

In Chapter 1 we provide an overview of the key concepts in hydrologic optics and aquatic remote sensing that provide the context for this study.

In Chapter 2 we provide a detailed evaluation of the uncertainty in downwelling irradiance estimations with diffuse reflectors, a topic overlooked despite large sets of data, particularly in inland waters, collected with this methodology. Our results help to shed light into uncertainty budgets of past measurements and provide basis for suggestion for an improved measurement protocol. Based on the magnitude of the uncertainty, we are also able to ascertain the suitability of this method for all uses of downwelling irradiance in hydrologic optics.

In Chapter 3 we provide another detailed evaluation related to errors in water-leaving radiance measurements with the on-water approach, a relatively recent method for field spectroscopy. We specifically evaluate the deployment setup used in this research, in order to provide corrections to our measurements. Additionally, we provide evaluations of the sensitivity of the error and simulations to parameters like backscattering ratio and the angular distribution of the downwelling radiance. We further provide evaluation of different deployment setups and suggest possible improvements for the method. 

In Chapter 4 we describe the full set of paired optical and biogeochemical data collected in Belgian inland and coastal waters. Extensive evaluations are provided to assert data consistency as a proxy for its accuracy. Of particular interest is the pairing of inherent and apparent optical data with particle size distribution, mineral fraction and metabarcoding, providing detailed composition information that can be used for devolvement of hyperspectral algorithms. The data is made freely available in a stable repository. 

In Chapter 5 we describe and validate a method to extract additional spectral information from a system of overlapping wavebands and apply this concept to the Operational Land Imager (OLI) onboard Landsat 8. We show that a virtual band, termed contra-band, with information on the yellow-orange spectra region covering the absorption peak of phycocyanin can be extracted from the panchromatic band of OLI. We provide extensive evaluations of the uncertainty and generality of the algorithm and of the independence of the retrieved information.

In Chapter 6 we revisit the problem of optical detection of Phaeocystis globosa blooms in the Belgian Coastal Zone (BCZ) by evaluating the origin of the signal detected by previously proposed hyperspectral algorithms. We show that the signal detected is not specific of Phaeocystis globosa even when considering the phytoplankton assemblage composition of the BCZ, due to synchronous phenology of other blooming species producing similar optical signals. We further uncover an apparent general pigmentation pattern associated with presence of chlorophyll c 3 . We conclude by providing a new algorithm to detect this pigmentation pattern and provide a regional ecological interpretation linked to Harmful Algal Blooms (HABs).

The research in those chapters is extensively supported with additional information in the form of theory, data and analysis provided in various Appendices. Lastly, in Chapter 7 we provide a synthesis of this research, discussing the relevance of our findings in the scope of hydrologic optics and aquatic remote sensing.}},
  author       = {{Castagna Mourão e Lima, Alexandre}},
  keywords     = {{Phaeocystis globosa,Hydrologic optics,Remote sensing,Aquatic optics,Inherent optical properties,Contra-bands,Radiative transfer}},
  language     = {{eng}},
  pages        = {{XXII, 348}},
  publisher    = {{Ghent University. Faculty of Sciences}},
  school       = {{Ghent University}},
  title        = {{Optical methods, uncertainties, and algorithms for remote sensing of coastal and inland water quality}},
  year         = {{2023}},
}