Satellite Ocean Colour: Current Status and Future Perspective

Abstract

Spectrally resolved water-leaving radiances (ocean colour) and inferred chlorophyll concentration are key to studying phytoplankton dynamics at seasonal and interannual scales, for a better understanding of the role of phytoplankton in marine biogeochemistry; the global carbon cycle; and the response of marine ecosystems to climate variability, change and feedback processes. Ocean colour data also have a critical role in operational observation systems monitoring coastal eutrophication, harmful algal blooms, and sediment plumes. The contiguous ocean-colour record reached 21 years in 2018; however, it is comprised of a number of one-off missions such that creating a consistent time-series of ocean-colour data requires merging of the individual sensors (including MERIS, Aqua-MODIS, SeaWiFS, VIIRS, and OLCI) with differing sensor characteristics, without introducing artefacts. By contrast, the next decade will see consistent observations from operational ocean colour series with sensors of similar design and with a replacement strategy. Also, by 2029 the record will start to be of sufficient duration to discriminate climate change impacts from natural variability, at least in some regions. This paper describes the current status and future prospects in the field of ocean colour focusing on large to medium resolution observations of oceans and coastal seas. It reviews the user requirements in terms of products and uncertainty characteristics and then describes features of current and future satellite ocean-colour sensors, both operational and innovative. The key role of in situ validation and calibration is highlighted as are ground segments that process the data received from the ocean-colour sensors and deliver analysis-ready products to end-users. Example applications of the ocean-colour data are presented, focusing on the climate data record and operational applications including water quality and assimilation into numerical models. Current capacity building and training activities pertinent to ocean colour are described and finally a summary of future perspectives is provided.

Keywords: capacity building; climate data records; ground-segment; ocean colour; phytoplankton; water-quality.

FIGURE 1 |

Screenshots of CyAN mobile app (Schaeffer et al., 2018): (A) The main page of the app allows for dropping locations of interest pins using latitude and longitude coordinates or by responsive touch on the screen. (B) The user is also able to set pin colour thresholds and an alert icon setting with the cogwheel based on their own criteria. (C) selecting a pin allows the user to visualise the complete satellite tile for the location of interest for spatial patterns, such as demonstrated with a coastal estuary on the west Florida shelf where Google maps is a transparent layer for quick geographic reference. (D) The user can view a temporal series of images related to their pin location and visualise temporal trends amongst different locations, where each line graph is a single pixel (not shown).

FIGURE 2 |

Screenshots for ESA ocean colour CCI OGC-compliant visualisation/analysis portal: (A) monthly version 3.1 global chl-a map (comprising merged bias-corrected SeaWiFS, Aqua-MODIS, MERIS and VIIRS) with uploaded shapefile for image data extraction. (B) Time series plot for region of interest; vertical bars show standard deviation and background grey shows range.

FIGURE 3 |

ESA Ocean Colour Climate Change Initiative version 3.1 monthly images for July 2003: (A) chl-a concentration, (B) RMSE, and (C) bias expressed as log10chl-a; comparisons of monthly relative error with GCOS uncertainty criteria for (D) R_rs443 nm and (E) chl-a for July 2003.

FIGURE 4 |

(A) Relationship between chl-a and the ENSO climate mode: note that the scale of chl-a anomalies is inverted. chl-a image is from an annual climatology. The monthly multivariate ENSO Index (MEI) was downloaded from the NOAA website (http://www.esrl.noaa.gov/); monthly plotted chlorophyll data were taken from the OC-CCI/CMEMS data set (modified from von Schuckmann et al., 2016); CMEMS north-east Atlantic region: (B) Time series (1997–2016) plot; (C) anomaly map for 2016 (with respect to 1997–2014 reference period) (re-worked from von Schuckmann et al., 2018).

FIGURE 5 |

(A,B) Map depicting the location of abalone farms on the South African coastline; lower panel shows satellite-derived chl-a concentration (in mg m-3) from the OLCI sensor for the Walker Bay area between the 29th of December 2016 and the 10th of February 2017. White areas represent either cloud or chl-a algorithm failure (Pitcher et al., 2019).

FIGURE 6 |

(A) Ocean Colour Monitor-derived chl-a concentration with overlaid SST contours used to find PFZ regions. (B) Example PFZ advisory issued by INCOIS, Hyderabad on a routine basis for the Orissa coast of India showing both potential fishing zones (noted as ‘Pfzlines’) as well as restricted fishing areas (‘Restricted_zones’). Such advisories are provided for the entire Indian coastline (Credit: INCOIS, Hyderabad; see https://incois.gov.in/MarineFisheries/PfzAdvisory).

FIGURE 7 |

(A) Geostationary Ocean Colour Imager radiance composite images and corresponding chl-a images from 09:25 to 16:25 on 13 August 2013: the chl-a profile line (AA′ ) and the hourly chl-a investigation points (1–3) are marked; (B) Hourly GOCI chl-a values at points 1–3; and (C) chl-a values along line AA′ at 09:25, 11:25, 15:25, and 16:25 local time.

FIGURE 8 |

Improved simulation of the plankton community structure by assimilating ocean-colour plankton functional type (PFT) products into a marine ecosystem model of the North East Atlantic. Column (A) PFTs fractions from the assimilative simulation (i.e., ratio of the PFT chlorophyll to the total chlorophyll); (B) PFTs fractions from ocean-colour data; and (C) percentage difference between the RMSD of the PFT assimilation (1-month forecasts) and of the reference model simulation without assimilation; RMSD is the root-mean-square-deviation between grid-point time series of model output and ocean-colour data. Figure elaborated from Figure 8 in Ciavatta et al. (2018), delivered by the Copernicus Marine Environment Monitoring Service TOSCA project.