Correction of Sensor Saturation Effects in MODIS Oceanic Particulate Inorganic Carbon
IEEE Transactions on Geoscience and Remote Sensing
Institute of Electrical and Electronics Engineers (IEEE)
Open access. This work is licensed under a Creative Commons Attribution 3.0 License. For more information, see http://creativecommons.org/licenses/by/3.0/
The highly reflective nature of high particulate inorganic carbon (PIC) from calcifying plankton, such as surface blooms of Emiliana-huxleyi in the latter stages of their life cycle, can cause the saturation of the MODerate resolution Imaging Spectrometer (MODIS) visible spectrum ocean colour bands. This saturation results in errors in the standard MODIS oceanic PIC product, resulting in the highest PIC levels being represented as cloud-like gaps (missing data) in daily level 2 data, and as either gaps or erroneously low PIC values in temporally averaged data (e.g. 8 day level 3 data). A method is described to correct this error and reconstruct the missing data in the ocean color band data, by regressing the 1 km spatial resolution ocean color bands against MODIS higher-resolution (500 m spatial resolution) bands with lower sensitivities. The method is applied to all North Atlantic MODIS data from 2002 to 2014. This shows the effect on mean PIC concentration over the whole North Atlantic to be less than 1% annually and 2% monthly, but with more significant regional effects, exceeding 10% in peak months in some coastal shelf regions. Effects are highly localised and tend to annually reoccur in similar geographical locations. Ignoring these missing data within intense blooms is likely to result in an underestimation of the influence that coccolithophores, and their changing distributions, are having on the North Atlantic carbon cycle. We see no evidence in this 12-year time series of a temporal poleward movement of these intense bloom events.
The authors thank the NASA Goddard Space Flight Centre, Ocean Biology Processing Group (NASA OB.DAAC) for supplying data for this study and the NERC Earth Observation Data Acquisition and Analysis Service (NEODAAS) for processing them. JS and PL were also supported by the UK Natural Environment Research Council (NERC) projects CArbon/Nutrient DYnamics and FLuxes Over Shelf Systems (CANDYFLOSS, contract NE/K002058/1) and the Radiatively Active Gases from the North Atlantic Region and Climate Change (RAGNARoCC, contract NE/K002511/1).
This is the author accepted manuscript. The final version is available from IEEE via the DOI in this record.
Published online 4 December 2017