In-water component retrieval over Case 2 water using Landsat 8: Initial results

Ocean color satellites (i.e. MODIS, SeaWiFS) are designed for studying large-scale water bodies and therefore perform well for retrieving water quality parameters at a global scale (e.g. MODIS Chlorophyll-a product). However, medium scale (i.e. hundred of meters) water bodies are not resolvable by these kinds of satellites due to their spatial resolution (i.e. 1000m). This work presents a methodology to address this problem using the Landsat 8 satellite for color producing agents (CPAs) (chlorophyll-a, sediments and colored dissolved organic matter (CDOM)) retrieval over inland and coastal waters (Case 2 waters). First, a model-based empirical line method (ELM) is used to correct the Landsat 8 imagery for atmospheric effects, having water reflectance pixels as output. Then, the concentration of these water pixels are estimated by a nonlinear optimization routine that searches in a look-up table (LUT) created with the in water radiative transfer model, Hydrolight.