for [13 years, MERIS for 10 years, and MODISA for [10 years), data quality
may not be as robust during the extended mission as in the early years. Of these
sensors, only MODIS (both Terra and Aqua) is operational, with data quality
showing some degradation. Thus, it is critical to plan continuity missions for the
future to establish seamless time-series observations.
NASA, ESA, and several other international agencies have been actively
planning for future ocean color missions, among which are two identical Sentinel3 ESA satellites as part of the Global Monitoring for Environment and Security
(GMES) programme and three NASA missions recommended by the Decadal
Survey for Earth Science (NRC 2007). The OLCI (Ocean Land Colour Imager)
instrument on board Sentinel-3, to be launched in late 2014, will serve as a
continuity mission of MERIS with enhanced performance in spectral resolution
and revisit frequency. The three recommended NASA missions are: (1) AerosolClouds-Ecosystems (ACE); (2) Geostationary Coastal and Air Pollution Events
(GEO-CAPE, Fishman et al. 2012); and (3) Hyperspectral Infrared Imager (HyspIRI). Each of these NASA missions has ocean color capability and a unique set
of science goals (NRC 2007). In particular, the hyperspectral sensors on these
missions will enable improved Chl retrievals in coastal waters (e.g., Hoogenboom
et al. 1998; Brando and Dekker 2003). Currently, the missions are under development, with launch dates tentatively scheduled for 2020 and beyond. Under the
auspices of the climate initiative, NASA is also planning a Pre-ACE (PACE)
mission for a 2019 launch (NASA 2010). These ocean color continuity missions
require continued efforts in calibration and algorithm development to ensure crosssensor consistency. Such a consistency is extremely critical when studying decadal-scale ocean changes, as demonstrated by two independent studies using
CZCS and SeaWiFS to study Chl changes between the 1970–1980s and
1990–2000s (Antoine et al. 2005; Gregg et al. 2005). More recently, multi-decadal
oscillations of phytoplankton abundance in the global ocean from the two ocean
color missions were found to be driven by climate variability (Martinez et al.
2009), even after allowing for different calibrations and algorithms used for the
two sensors. SeaWiFS and MODIS/A operated simultaneously for a period of
years (2002–2010), and though both were managed by NASA, there are discrepancies between the Chl records. Recently, the CI algorithm has been shown to
yield more consistent Chl data records for the open ocean (Fig. 7.11, Hu et al.
2012b), where discrepancy between SeaWiFS and MODIS/A Chl records was
reduced by at least 50 % for most of the time period for global clear waters. Such
an improvement is expected to lead to reduced uncertainties in the globally merged
data products from multiple sensors (e.g., Maritorena et al. 2010).
While all sensor characterization and calibration procedures as well as algorithm development are mature for the open ocean, future emphasis will be on Chl
algorithm improvement for coastal waters. The effects of shallow bottom, CDOM,
and suspended sediments must be resolved either implicitly through empirical
regression or explicitly through semi-analytical inversion. Currently, there are
several approaches to use the red-NIR wavelengths to avoid the effect of CDOM,
and a preliminary approach to remove sediment effects empirically (Wynne et al.
192
C. Hu and J. Campbell
may not be as robust during the extended mission as in the early years. Of these
sensors, only MODIS (both Terra and Aqua) is operational, with data quality
showing some degradation. Thus, it is critical to plan continuity missions for the
future to establish seamless time-series observations.
NASA, ESA, and several other international agencies have been actively
planning for future ocean color missions, among which are two identical Sentinel3 ESA satellites as part of the Global Monitoring for Environment and Security
(GMES) programme and three NASA missions recommended by the Decadal
Survey for Earth Science (NRC 2007). The OLCI (Ocean Land Colour Imager)
instrument on board Sentinel-3, to be launched in late 2014, will serve as a
continuity mission of MERIS with enhanced performance in spectral resolution
and revisit frequency. The three recommended NASA missions are: (1) AerosolClouds-Ecosystems (ACE); (2) Geostationary Coastal and Air Pollution Events
(GEO-CAPE, Fishman et al. 2012); and (3) Hyperspectral Infrared Imager (HyspIRI). Each of these NASA missions has ocean color capability and a unique set
of science goals (NRC 2007). In particular, the hyperspectral sensors on these
missions will enable improved Chl retrievals in coastal waters (e.g., Hoogenboom
et al. 1998; Brando and Dekker 2003). Currently, the missions are under development, with launch dates tentatively scheduled for 2020 and beyond. Under the
auspices of the climate initiative, NASA is also planning a Pre-ACE (PACE)
mission for a 2019 launch (NASA 2010). These ocean color continuity missions
require continued efforts in calibration and algorithm development to ensure crosssensor consistency. Such a consistency is extremely critical when studying decadal-scale ocean changes, as demonstrated by two independent studies using
CZCS and SeaWiFS to study Chl changes between the 1970–1980s and
1990–2000s (Antoine et al. 2005; Gregg et al. 2005). More recently, multi-decadal
oscillations of phytoplankton abundance in the global ocean from the two ocean
color missions were found to be driven by climate variability (Martinez et al.
2009), even after allowing for different calibrations and algorithms used for the
two sensors. SeaWiFS and MODIS/A operated simultaneously for a period of
years (2002–2010), and though both were managed by NASA, there are discrepancies between the Chl records. Recently, the CI algorithm has been shown to
yield more consistent Chl data records for the open ocean (Fig. 7.11, Hu et al.
2012b), where discrepancy between SeaWiFS and MODIS/A Chl records was
reduced by at least 50 % for most of the time period for global clear waters. Such
an improvement is expected to lead to reduced uncertainties in the globally merged
data products from multiple sensors (e.g., Maritorena et al. 2010).
While all sensor characterization and calibration procedures as well as algorithm development are mature for the open ocean, future emphasis will be on Chl
algorithm improvement for coastal waters. The effects of shallow bottom, CDOM,
and suspended sediments must be resolved either implicitly through empirical
regression or explicitly through semi-analytical inversion. Currently, there are
several approaches to use the red-NIR wavelengths to avoid the effect of CDOM,
and a preliminary approach to remove sediment effects empirically (Wynne et al.
192
C. Hu and J. Campbell