3 Thermal Infrared Remote Sensing and Sea Surface Temperature . . .
57
additional bands in the visible and near-infrared parts of the spectrum for specific
purposes, complementing information given by the far-infrared bands.
The strength of the infrared radiation emitted by the ocean is a function of the
temperature at its surface: the higher the temperature, the greater is the radiant
energy. Sea surface temperature is therefore the main product retrieved by these
sensors, after correcting the water surface-emitted signal from the contribution due
to the atmosphere. Note that a complete description of the retrieval of SST in terms
of the physics of the signal and atmospheric effects is outside the scope of this paper.
Details on these issues are available through a number of technical publications
(e.g. Kilpatrick et al. 2001; Merchant et al. 2008) and books (Robinson 2004 and
references therein).
The reliable calibration of the instruments, coupled with the regular validation
exercises with field observations, allow SST measurements from TIR sensors to
reach high levels of absolute accuracy, e.g. < 0.2
◦ C with the Advanced AlongTrack Scanning Radiometer (AATSR) instrument (O’Carrol et al. 2008). In addition,
polar-orbiting satellites can provide SST measurements twice daily at a spatial resolution of ca. 1 km. Alternatively, radiometers on-board geostationary satellites—e.g.
the Meteosat Second Generation (MSG) satellites of the European Space Agency
(ESA)—have coarser ground resolution (3–5 km) but much higher temporal sampling frequency (every 15 min or half-hourly) within its large but constant field of
view.
Thermal sensors operating onboard satellite have been in orbit for the last 30 years,
providing multi-decadal observations of the marine and ocean dynamics through SST
measurements at global and regional scales. The Advanced Very High Resolution
Radiometer (AVHRR) series operated by the US National Oceanic and Atmospheric
Administration (NOAA) National Environmental Satellite, Data, and Information
Service (NESDIS) represent undoubtedly the longest time series of SST measurements collected from space. The data are distributed operationally and globally
through the AVHRR Pathfinder programme
1 (Kilpatrick et al. 2001) from 1981 to
present, at spatial resolution of 4 km and daily daytime and nighttime fields. The
processing and validation of Pathfinder SST are outside the scope of this work, but
have been the focus of several publications (Kearns et al. 2000; Kumar et al. 2003;
Marullo et al. 2007; Nykjaer 2009).
In addition to AVHRR, many other satellite missions equipped with thermal infrared bands to measure SST have been launched since 1999 in a variety of orbits
(Donlon 2010). Such a multitude of available SST datasets derived from sensors
with different characteristics and processed in slightly different ways called for an
international effort to harmonize these data in order to provide users with consistent
long-term data stream of SST products. As part of Global Ocean Data Assimilation
Experiment (GODAE), the Group for High Resolution SST (GHRSST) provides a
wide variety of SST data
2 (Donlon et al. 2007), including a multi-sensor merged
product and uncertainty estimates for each of the products (e.g. Level 4 K10 SST
1 The link http://www.nodc.noaa.gov/SatelliteData/pathfinder4km provides access to Pathfinder
through the present, adding observations by the NOAA-19 sensor.
2 Available at https://www.ghrsst.org.
57
additional bands in the visible and near-infrared parts of the spectrum for specific
purposes, complementing information given by the far-infrared bands.
The strength of the infrared radiation emitted by the ocean is a function of the
temperature at its surface: the higher the temperature, the greater is the radiant
energy. Sea surface temperature is therefore the main product retrieved by these
sensors, after correcting the water surface-emitted signal from the contribution due
to the atmosphere. Note that a complete description of the retrieval of SST in terms
of the physics of the signal and atmospheric effects is outside the scope of this paper.
Details on these issues are available through a number of technical publications
(e.g. Kilpatrick et al. 2001; Merchant et al. 2008) and books (Robinson 2004 and
references therein).
The reliable calibration of the instruments, coupled with the regular validation
exercises with field observations, allow SST measurements from TIR sensors to
reach high levels of absolute accuracy, e.g. < 0.2
◦ C with the Advanced AlongTrack Scanning Radiometer (AATSR) instrument (O’Carrol et al. 2008). In addition,
polar-orbiting satellites can provide SST measurements twice daily at a spatial resolution of ca. 1 km. Alternatively, radiometers on-board geostationary satellites—e.g.
the Meteosat Second Generation (MSG) satellites of the European Space Agency
(ESA)—have coarser ground resolution (3–5 km) but much higher temporal sampling frequency (every 15 min or half-hourly) within its large but constant field of
view.
Thermal sensors operating onboard satellite have been in orbit for the last 30 years,
providing multi-decadal observations of the marine and ocean dynamics through SST
measurements at global and regional scales. The Advanced Very High Resolution
Radiometer (AVHRR) series operated by the US National Oceanic and Atmospheric
Administration (NOAA) National Environmental Satellite, Data, and Information
Service (NESDIS) represent undoubtedly the longest time series of SST measurements collected from space. The data are distributed operationally and globally
through the AVHRR Pathfinder programme
1 (Kilpatrick et al. 2001) from 1981 to
present, at spatial resolution of 4 km and daily daytime and nighttime fields. The
processing and validation of Pathfinder SST are outside the scope of this work, but
have been the focus of several publications (Kearns et al. 2000; Kumar et al. 2003;
Marullo et al. 2007; Nykjaer 2009).
In addition to AVHRR, many other satellite missions equipped with thermal infrared bands to measure SST have been launched since 1999 in a variety of orbits
(Donlon 2010). Such a multitude of available SST datasets derived from sensors
with different characteristics and processed in slightly different ways called for an
international effort to harmonize these data in order to provide users with consistent
long-term data stream of SST products. As part of Global Ocean Data Assimilation
Experiment (GODAE), the Group for High Resolution SST (GHRSST) provides a
wide variety of SST data
2 (Donlon et al. 2007), including a multi-sensor merged
product and uncertainty estimates for each of the products (e.g. Level 4 K10 SST
1 The link http://www.nodc.noaa.gov/SatelliteData/pathfinder4km provides access to Pathfinder
through the present, adding observations by the NOAA-19 sensor.
2 Available at https://www.ghrsst.org.