Computation of primary production from remote sensing of ocean colour at the northwestern Atlantic C-SOLAS Lagrangian site

摘要

ABSTRACT: At the C-SOLAS Lagrangian Study site, the phytoplankton community was dominated by diatoms in declining bloom conditions, which were characterised by a decrease in primary production over time. Measurements of primary production were performed using photosynthesis-light experiments and simulated in situ incubations. Several methods were used to assign photosynthetic parameters for estimation of primary production from remote-sensing of ocean colour: (1) the nearest-neighbour method (NNM), (2) a temperature-dependent model, (3) an iterative approach to retrieve photosynthetic parameters from in situ measurements of phytoplankton production, and (4) average values of measured parameters. Owing to the declining status of the diatom population and its patchy distribution, the magnitude of primary production measured from in situ incubations was highly variable. Under bloom conditions, all methods underestimated production compared with results of simulated in situ measurements; however, the NNM provided the closest estimates to the in situ measurements. In the declining bloom phase, the NNM overestimated primary production, whereas the iterative method returned estimates of production in good agreement with the in situ observations. In conditions characterized by mixed-phytoplankton populations, the NNM returned estimates in agreement with the observations. A new method is proposed for classifying image pixels using criteria accessible to remote sensing, according to the phase of the diatom bloom (bloom versus declining conditions) and according to community composition (diatom versus mixed population). Upon separating the database used for parameter assignment into 2 parts, one representing the general phytoplankton community and the other representing declining diatom conditions, estimates of chlorophyll a-normalised water-column production were in good agreement with in situ observations.