Tools and algorithms for drone and satellite based ocean color science
Currently in the process of implementing a few of the NASA Ocean Color team's chlorophyll retrieval algorithms from https://oceancolor.gsfc.nasa.gov/atbd/chlor_a/. The goal is to make these algorithms more easily understandable and to allow fine-tuning for sensors that differ from the primary NASA ocean observing satellites, particularly for drone-based multispectral sensors. This is all done right now using the NASA bio-Optical Marine Algorithm Dataset (NOMAD) available at https://seabass.gsfc.nasa.gov/wiki/NOMAD.
Currently implemented algorithms:
- Hu et al 2012 color index (https://doi.org/10.1029/2011JC007395) - code
- complete and exactly replicates the nonlinear regression coefficients, r-squared, and RMS from the paper.
- this is the algorithm NASA uses for oligotrophic (low-productivity) waters in the global ocean
- OCx algorithm (originally via https://doi.org/10.1029/98JC02160 but more recently updated on https://oceancolor.gsfc.nasa.gov/atbd/chlor_a/) - code
- nearly exact but the exact filtering procedure and curve matching approach isn't outlined
- will be updating to match https://doi.org/10.1016/j.rse.2019.04.021 at some point
Application examples
- applying the OCx algorithm on Landsat 8's Provisional Aquatic Reflectance product - code
January 5, 2021 image from VIIRS (credit NASA Ocean Color)