Your access to IMOS Satellite Remote Sensed SST data discovery and exploration is through the Australian Ocean Data Network (AODN) Portal.
The highest spatial resolution (1.1 km x 1.1 km) data from Advanced Very High Resolution Radiometer (AVHRR) sensors on NOAA Polar-orbiting Operational Environmental Satellites (POES) can only be obtained through receiving direct broadcast High Resolution Picture Transmission (HRPT) data from the satellite. The Bureau of Meteorology, in collaboration with CSIRO Oceans and Atmosphere Flagship, combines raw HRPT data from reception stations located in Australia and Antarctica and produces real-time, level 2 (geolocated, swath) and level 3 (gridded, composite) files containing either “skin” (~ 10 micron depth) or “foundation” (~ 10 m depth) SSTs. All files follow the latest International Group for High Resolution Sea Surface Temperature (GHRSST) “GDS 2” file formats. The IMOS SST processing system aims to provide a consistent, accurate record of SST over an extended period of time from AVHRR sensors on all operational NOAA POES satellites from NOAA-11 to NOAA-19. The SST values are derived by regressing the brightness temperature observations from the satellite sensor against collocated drifting buoy SST observations. The system smoothly and automatically corrects for changes in each sensor’s bias using a running one year calibration window, adjusted monthly (Griffin et al., 2017).
There are several advantages to using the IMOS-GHRSST format SST products. Quality level flags and error information enable users to tailor the SST data for their particular research application by trading spatial coverage for accuracy and vice versa. Users have the ability to access the GHRSST-format gridded SST products through IMOS OPeNDAP servers, greatly simplifying data access and extraction. In addition, each GHRSST file netCDF header contains very comprehensive metadata information.
For each SST value the GHRSST files contain a quality level flag (“quality_level”), bias and standard deviation error estimates (“sses_bias” and “sses_standard_deviation”) and flags (“l2p_flag”). The quality level (0, 1, …, 5) is based solely on the SST pixel’s proximity to cloud in kilometres, and is therefore a measure of the amount of cloud contamination expected. Best quality pixels (quality_level = 5) are at least 5 km horizontally from the nearest detected cloud. In the IMOS SST processing system, the bias and standard deviation estimates for each SST value are a smoothly varying function of satellite zenith angle, time of day, geographic location and quality level, calculated using a rolling one year window of match-ups with drifting buoy SST data, adjusted approximately every 5 days. The l2p_flags can be used to filter the data for land, high aerosol content, high or low winds, etc, and are described in each file header.
The following IMOS GHRSST “GDS 2” format files are produced in netCDF4 format using the Australian HRPT AVHRR SST data over the Australian domain (70°E to 190°E, 70°S to 20°N) and have been reprocessed back to 1992:
- L2P: Single swath, geolocated SSTskin (used in the Bureau of Meteorology’s operational daily SST analyses)
- L3U: Single swath, gridded SSTskin at 0.02° x 0.02° resolution (used in IMOS OceanCurrent and FishTrack)
- L3C: Single sensor, multiple swath, gridded SSTskin at 0.02° x 0.02° resolution, averaged over 1 and 3 days (used for diurnal warming research, e.g. TWP+ Project)
- L3S: Multiple sensors, multiple swath, day-only or night-only SSTskin and day+night foundation SST (SSTfnd), gridded at 0.02° x 0.02° resolution and averaged over multiple days (1, 3, 6, 14 and 1 month) (used for ReefTemp NextGen, IMOS OceanCurrent and marine biological research).
The following two graphics explain the various IMOS GHRSST products.
Examples of the various different types of Australian domain IMOS GHRSST-format products from locally received AVHRR SST data.
A graphical representation of how different day or night swath data are combined from multiple satellites to produce the various GHRSST SST products (L2P, L3U, L3C and L3S).
Multiple swath L3C and L3S products produced using only day-time data are stored in the “day” directories, night-time data in the “night” directories and day and night data in the “dn” directories. Please note that the ghrsst_fv01 directory contains the “fv01” version of the IMOS HRPT AVHRR SST products as documented in Paltoglou et al. (2010) and reprocessed back to 1998 using data from NOAA-15 to NOAA-19 satellites. These files are currently available up to real-time. The ghrsst_fv02 directory contains the fv02 version of the products as documented in Beggs et al. (2013) and Griffin et al. (2017). The fv02 products have been reprocessed back to 1992 using data from NOAA-11 to NOAA-19 satellites. The fv02 files represent a more complete data record than fv01 and use different algorithms to convert from AVHRR brightness temperatures to SST values (Griffin et al., 2017). It is expected that the fv02 products will be produced operationally in real-time at the Bureau of Meteorology in 2015. The ghrsst directory contains all fv02 files up to 31 December 2014 and fv01 files after that date.
Southern Ocean domain (2.5°E to 202.5°E, 77.5°S to 27.5°S) 0.02° x 0.02° resolution gridded fv01 and fv02 L3U and L3S AVHRR SST products are also available from the ghrsst, ghrsst_fv01 and ghrsst_fv02 directories. Southern Ocean domain directories can be distinguished from Australian domain by an appended ‘S’ in the directory name. For example, Australian domain 1-day L3S files can be found at rs-data1-mel.csiro.au/thredds/catalog/imos-srs/sst/ghrsst/L3S-1d/catalog.html and Southern Ocean domain 1-day L3S files can be found at rs-data1-mel.csiro.au/thredds/catalog/imos-srs/sst/ghrsst/L3S-1dS/catalog.html.
Example of a Southern Ocean domain IMOS GHRSST-format product
Some additional products (SST and Ocean Colour) are available through the THREDDS catalogue maintained by CSIRO.
The Satellite Remote Sensing Facility Publication Report - If you have any questions regarding the data, or corrections, or would like to add a publication or presentation that uses IMOS data please contact the IMOS office via email: publication(at)emii.org.au.