258
H.M. Beggs
Some newer experimental analyses such as the 6-hourly, 4 km resolution,
SSTblend Analysis over the Gulf of Maine, US (Vinogradova et al., 2009) combine
hourly IR SST from geostationary satellites with daily multi-sensor SST analyses
to attempt to resolve the diurnal cycle. Other methods being trialled for operational implementation in Numerical Weather Prediction (NWP) data assimilation
systems (see Section 15.7) have been hourly skin SST analyses derived from foundation SST analyses combined with hourly SSTskin – SSTfnd estimates (Beggs
et al., 2009a). These SSTskin – SSTfnd values can be calculated in real-time by
inputting NWP forecast surface winds into an empirical model based on geostationary SEVIRI satellite IR SSTskin observations and AMSR-E wind observations
(Gentemann et al., 2003).
15.4.1 TIR Requirements
It is very useful for operational SST analysis (level 4) producers to be able to obtain
level 2 satellite SST products in a common format with associated error estimates
and quality flags. The GHRSST L2P format products (Section 15.2) are very effective as they contain all the variables in one file required to use the SST data in an OI
analysis system. These include error estimates (bias and standard deviation) for each
SST value along with quality control and cloud proximity flags, surface wind and
surface solar insolation (from either observations or NWP models), aerosol optical
depth and time of observation.
Most of the operational SST analysis systems listed in Table 15.4 use some or
solely GHRSST L2P IR SST products and several (OSTIA, ODYSSEA, NCODA,
GAMSSA, AVHRR_OI and AVHRR_AMSR_OI) are available in the GHRSST-L4
netCDF format incorporating useful ancillary data fields such as analysis error, sea
ice concentration and land mask.
With the future operational implementation of current experimental ultra-highresolution SST analyses such as G1SST, there will be an increasing requirement
from operational SST analysis systems for timely, accurate, 1 km resolution, IR
SST level 2 products with reliable error statistics and effective cloud masking.
15.5 Ocean Models
Operational numerical ocean models providing forecasts of currents, temperature and salinity fields are used for a variety of applications including coral reef
management, tide predictions, marine sanctuary and estuary management, diving
operations, naval applications, oil and chemical spill drift forecasts, search and rescue operations, offshore oil drilling operations, cable laying and ship routing. Today,
a dozen numerical ocean modelling systems routinely operate in the nine countries that participated in the Global Ocean Data Assimilation Experiment (GODAE)
(Dombrowsky et al., 2009). They range from regional high resolution systems that
H.M. Beggs
Some newer experimental analyses such as the 6-hourly, 4 km resolution,
SSTblend Analysis over the Gulf of Maine, US (Vinogradova et al., 2009) combine
hourly IR SST from geostationary satellites with daily multi-sensor SST analyses
to attempt to resolve the diurnal cycle. Other methods being trialled for operational implementation in Numerical Weather Prediction (NWP) data assimilation
systems (see Section 15.7) have been hourly skin SST analyses derived from foundation SST analyses combined with hourly SSTskin – SSTfnd estimates (Beggs
et al., 2009a). These SSTskin – SSTfnd values can be calculated in real-time by
inputting NWP forecast surface winds into an empirical model based on geostationary SEVIRI satellite IR SSTskin observations and AMSR-E wind observations
(Gentemann et al., 2003).
15.4.1 TIR Requirements
It is very useful for operational SST analysis (level 4) producers to be able to obtain
level 2 satellite SST products in a common format with associated error estimates
and quality flags. The GHRSST L2P format products (Section 15.2) are very effective as they contain all the variables in one file required to use the SST data in an OI
analysis system. These include error estimates (bias and standard deviation) for each
SST value along with quality control and cloud proximity flags, surface wind and
surface solar insolation (from either observations or NWP models), aerosol optical
depth and time of observation.
Most of the operational SST analysis systems listed in Table 15.4 use some or
solely GHRSST L2P IR SST products and several (OSTIA, ODYSSEA, NCODA,
GAMSSA, AVHRR_OI and AVHRR_AMSR_OI) are available in the GHRSST-L4
netCDF format incorporating useful ancillary data fields such as analysis error, sea
ice concentration and land mask.
With the future operational implementation of current experimental ultra-highresolution SST analyses such as G1SST, there will be an increasing requirement
from operational SST analysis systems for timely, accurate, 1 km resolution, IR
SST level 2 products with reliable error statistics and effective cloud masking.
15.5 Ocean Models
Operational numerical ocean models providing forecasts of currents, temperature and salinity fields are used for a variety of applications including coral reef
management, tide predictions, marine sanctuary and estuary management, diving
operations, naval applications, oil and chemical spill drift forecasts, search and rescue operations, offshore oil drilling operations, cable laying and ship routing. Today,
a dozen numerical ocean modelling systems routinely operate in the nine countries that participated in the Global Ocean Data Assimilation Experiment (GODAE)
(Dombrowsky et al., 2009). They range from regional high resolution systems that