16:45–17:00 | ABSTRACT 1225
HIGH PERFORMANCE OF PHYTOPLANKTON PIGMENT ESTIMATION IN THE ATLANTIC OCEAN FROM HYPERSPECTRAL ABOVE-WATER REFLECTANCE
Estimation of phytoplankton pigments from ocean colour radiometry is a key step toward characterisation of phytoplankton functional groups at the global scale. Although this is a long-standing area of research, the accuracy and number of pigments retrieved from multispectral satellites have been limited. Hyperspectral remote sensing (e.g. from NASA PACE and future ESA CHIME) is expected to provide a step-change in the performance of pigment retrieval algorithms. Development and testing of these new algorithms ideally require co-located measurements of hyperspectral reflectance and phytoplankton pigment concentrations. To this end, we present an eight-cruise compilation of hyperspectral above-water reflectance from the UK Atlantic Meridional Transect (AMT) co-located with > 200 surface HPLC pigment samples. The near-continuous reflectance data from the AMT provides proof-of-concept for mapping pigments along cruise tracks and use of autonomous data for algorithm development. We use the AMT database to train, tune, and evaluate the performance of a hyperspectral algorithm that can potentially provide global pigment products from PACE. The algorithm uses bio-optical modelling and spectral derivatives to isolate higher-frequency absorption features, which are used to construct a statistical model for pigment concentrations. We demonstrate high performance (absolute percentage error ~ 15-35%, log correlation coefficients > 0.75) for the estimation of > 15 pigments. We compare with Chlorophyll-a co-variation relationships, which illustrates the additional information that can be obtained from hyperspectral data. We conclude by describing how the AMT data are being combined with other high-quality measurements to produce an extensive hyperspectral algorithm development dataset across a range of water types.
Thomas Jordan, Plymouth Marine Laboratory, [email protected], https://orcid.org/0000-0002-2096-8858
Gavin Tilstone, Plymouth Marine Laboratory, [email protected]
Junfang Lin, Plymouth Marine Laboratory, [email protected]
Robert Brewin, University of Exeter, [email protected]
Giorgio Dall’olmo, Istituto Nazionale di Oceanografia e di Geofisica Sperimentale, [email protected]
Federico Ienna, Istituto Nazionale di Oceanografia e di Geofisica Sperimentale, [email protected]
Carlos Rafael Mendes, University of Lisbon and Federal University of Rio Grande, [email protected]
Raul Costa, University of Southampton and Federal University of Rio Grande, [email protected]
Crystal Thomas, NASA Goddard Space Flight Center, [email protected]
Maria Laura, Zoffoli, Consiglio Nazionale delle Ricerche, Istituto di Scienze Marine (Trieste), [email protected]
Emanuele Organelli, Consiglio Nazionale delle Ricerche, Istituto di Scienze Marine (Rome), [email protected]
Victor Martinez-Vicente, Plymouth Marine Laboratory, [email protected]
