6. Inverse Modeling and Validation

The ‘inverse’ model solves the IOPs (or concentrations of optically active water constituents) from reflectance measured at the top of the atmosphere by remote sensors or from in situ measurements of reflectance. Inversion approaches for the retrieval of water-quality parameters from remotely sensed data are typically divided into two categories—empirical and model based (IOCCG 2002). Empirical methods derive a relationship between optical measurements and constituent concentrations based on observations from experimental data sets. Model-based techniques exploit bio-optical models that capture the connection between water constituent and spectra of water-leaving radiance/reflectance and a radiative transfer model that describes light propagation through the water and the atmosphere. Model-based solutions have the ability to retrieve simultaneously various optical and biogeophysical parameters from the remote-sensing reflectance. Their theoretical basis on sound solutions to the radiative transfer equation makes them potentially a more expedient approach (Matthews 2011).