GHRSST fv02 file fields

Description of IMOS GHRSST fv02 file fields

sea_surface_temperature: The SST based on a non-linear retrieval scheme with separate day and night algorithms (Griffin et al., in prep.). The retrieval scheme is calculated monthly on a two year rolling window regression of observations against in situ measurements.

sea_surface_temperature_day_night: An optional SST based on a retrieval scheme which uses a single algorithm for both day and night processing (Griffin et al., in prep.), resulting in a consistent retrieval for both day and night. The SST thus retrieved is thus expected to be more suited to studies of the diurnal cycle. It is expected that when there is a large deviation between sea_surface_temperature and sea_surface_temperature_day_night during the day, that this corresponds to a transitory atmospheric condition. Large differences thus detected, of 3sses_standard_deviation or greater, result in a flag to be set in the l2p_flags field. The retrieval scheme is calculated monthly on a 2 year rolling window regression of observations against in situ measurements.

quality level: The pixel quality based on a measure of the distance to the nearest cloud in kilometers, or 5 if the distance is greater than 5 km. Pixel quality is assumed to be primarily driven by the degree of cloud contamination, so this forms the best first indication of the quality of a given pixel. Further indications of pixel quality can be garnered from l2p_flags, ancillary information, and the sses_bias and sses_standard_deviation which provide estimates of bias and error based on in situ measurements under similar view, quality and geographical conditions.

l2p_flags: Additional l2p_flags have been added to reflect various ancillary conditions. This allows decisions on data quality related to ancillary fields to be made by reading a bit array rather than processing the ancillary fields.

sses_count: A measure of the indicative number of degrees of measurements which went into validating or generating the SST measurement. For L2P files, this is an indication of the number of in situ measurements made under similar view, quality and geographical conditions. For L3U files, this is an indication of the number and proportion of L2P observed pixels that went into the composition of the gridded cell. For L3C and L3S files, this is an indication of the weighted number of degrees of freedom that contributed to the SST. Empirical degrees of freedom models are updated every five days and represent relatively high frequency changes in the condition of the SST retrieval.

sses_bias: An estimate of the bias of the measurement with respect to in situ measurements made under similar view, quality and geographical conditions. The sses_bias is empirically determined, and not based on regression with physically inspired SST retrieval models. However, addition of the bias does result in an overall more accurate match between retrieved SST and in situ measurements. sses_bias can be subtracted from the SST if this empirical compensation is required to be included in the assessment of the SST. sses_bias strictly is derived from sea_surface_temperature, and should not be blindly applied to sea_surface_temperature_day_night.  Empirical bias models are updated every five days and represent relatively high frequency changes in the condition of the SST retrieval.

sses_standard_deviation: An indication of the uncertainty associated with the SST estimate, compared to in situ measurements made under similar view, quality and geographical conditions.  The sses_standard_deviation is empirically determined, and not based on regression with physically inspired SST retrieval models. However, sses_standard_deviation can be compared to the sses_bias to give some idea about the confidence in using this as a correction to the SST.  sses_standard_deviation strictly is derived from sea_surface_temperature, and thus will not directly apply to sea_surface_temperature_day_night, however, it may be used to determine the scale on which the difference between the two sea surface temperatures may show significant difference. Empirical standard deviation models are updated every five days and represent relatively high frequency changes in the condition of the SST retrieval.

sst_count: The unbiased number of SST observations that were merged into the pixel at this location.  This field is not present in L2P files, and will be absent if all valid sst_count values are 1. This field is typically used in multi-day L3S over long periods of time, to correct SSES estimates.

sst_mean: The unbiased mean of the SST observations that were merged into the pixel. This field is not present in L2P files, and will be absent if sst_count is less than 2. This field is typically used in multiday L3S over long periods of time, to correct SSES estimates.

sst_standard_deviation: The unbiased standard deviation of the SST observations that were merged into the pixel. This field is not present in L2P files, and will be absent if all valid values are ill-defined because sst_count is less than 2. This field is typically used in multi-day L3S over long periods of time, to correct SSES estimates.

dt_analysis: The deviation between the SST sea_surface_temperature, and a level 4 analysis of the previous day's foundation SST observations. For skin SST, the mean value of dt_analysis in a wind environment that encourages top level mixing, will be 0.17K. Prior to 23 July 2008, dt_analysis makes use of the NCDC SSTblend analysis (Reynolds et al., 2007). After 23 July 2008, GAMSSA foundation SST is used as the level 4 analysis (Zhong and Beggs, 2008).

wind_speed: The surface wind estimated from global numerical weather prediction models and interpolated in time and space to the location of the SST value.  Prior to 1 September 2009, ECMWF historical re-analysis is used. After 1 September 2009, the Australian Bureau of Meteorology ACCESS-G analysis or forecast 10 m surface winds are used (Puri et al., 2010).  Note that the model wind speeds are available on a much coarser spatial and temporal resolution than the SST observations and are included in the IMOS-GHRSST files for quality control purposes.

sea_ice_fraction: A number between zero and one, representing the sea-ice fraction, is taken from the NCEP sea ice analysis of the previous day.

aerosol_dynamic_indicator: After 27 November 1998 the Aerosol Dynamic Indicator (ADI) is added based on the NOAA AERO100 data set analysis based on the previous day. OS-DPD daily files are used after 11 January 2011. Early SST retrievals prior to 27 November 1998, may not contain ADI fields.

history: The history metadata contains much information about the history of the file during processing as well as the tools that processed the file and other status information that relates to the processing.