Page 440 - Economia Azul - versão: inglês

This project involved the largest operation- OceanPredict, new challenges are now time, always employing the Hybrid Coordi- 30 days, the errors of the SST predictions in
al oceanography centers in the world and being discussed, in particular how opera- nate Ocean Model (HYCOM) (MELLO et al., relation to the analyses reached an average
supported the planning and carrying out of tional oceanography will support the de- 2013; TANAJURA, et al., 2013). Today, a of 0.5 ºC and, without assimilation, around
scientific studies involving the improvement velopments of the United Nations Decade grid with 1/12 of horizontal resolution and 1 ºC. The SSHA correlations of the pre-
o
of in situ and sensing observation systems, of Ocean Science Program for Sustainable 32 vertical layers is used over the Atlantic dictions initialized with assimilation were
oceanographic data assimilation methods Development and all the services to society Ocean and part of the Indian Ocean (60 W about 0.8, and, without assimilation, less
o
and ocean models (CUMMINGS et al., and the Blue Economy that are expected. – 40 E, 78 S – 50 N) and a grid with 1/24 than 0.6. Other results on the predictabil-
o
o
o
o
2009; DAVIDSON et al., 2009). In 2009, REMO: first operational numerical and 32 layers for Metarea V that considers ity of HYCOM+RODAS in a window of up
GODAE was terminated, and its legacy was prediction system with assimilation in the effects of tides. The system is initialized to 3 days were accessed in Tanajura et al.
passed on to GODAE OceanView (BELL et Latin America by the REMO Ocean Data Assimilation Sys- (2020) and Santana et al. (2020). Reduc-
al., 2015). In this new stage, the same gen- tem (RODAS) (TANAJURA et al., 2014; MI- tions in SST prediction errors were greater
eral objectives of GODAE were maintained, In Brazil, a major advance in operational GNAC et al., 2015; TANAJURA et al., 2020, than 50% and increases in SSH correlation
but the work was adapted to meet new de- oceanography was made in 2007 with the SANTANA et al., 2020; DORFSCHÄFER et greater than 40%. These examples demon-
mands for the development of operational creation of the Oceanographic Modeling al., 2020), based on the ensemble optimal strate the great importance of observations
coastal oceanography, regional forecasting and Observation Network (REMO), with interpolation method (EnOI), very suitable and assimilation in the predictive system.
systems with high spatial resolution and resources from Petrobras and the Nation- for operational use in view of its relatively In addition to being used in forecasts,
better dissemination of observational and al Agency of Petroleum, Natural Gas and low computational cost (EVENSEN, 2003; objective analyses can positively contribute
forecasting products or simulation. Still, Biofuels (ANP) (LIMA et al., 2013; TANA- OKE and SCHILLER, 2007; COUNILLON to a better understanding of physical and/
scientific challenges are permanently faced JURA et al., 2013; MELLO et al., 2013). A and BERTINO, 2009). The HYCOM+RODAS or biogeochemical processes in the oceans,
with the implementation of new monitor- consortium comprising UFBA, UFRJ, FURG, operating at the CHM produces daily fore- since, as they are more accurate than pure-
ing systems offering new data, such as the USP, IEAPM, the Centro de Hidrografia da casts in the 5-day window for the entire At- ly model fields, they can also offer more ac-
salinity of the sea surface obtained by the Marinha (CHM) and CENPES/Petrobras lantic Ocean forced by atmospheric fields curate diagnoses of the circulation and the
Soil Moisture and Ocean Salinity (SMOS) supported the development in ocean mod- predicted by the Global Forecasting System variability of the oceans. Once the quality of
satellite from 2009 (REUL et al., 2014), eling and data assimilation, with the CHM (GFS) model of the National Centers for En- the analyses is proven, they can also serve as
that can and should be assimilated in or- as the operational center. Today, the insti- vironmental Prediction (NCEP) of NOAA on a complement to observational monitoring
der to improve the initial condition of the tutions that form REMO are UFBA, UFRJ, the 1/12 grid and by the COSMO model and a reference for evaluating the quality
0
forecasting systems (MARTIN et al., 2016; CENPES and CHM, but new collaborations integrated into the CHM in the 1/24 grid. of simulations and predictions. In addition,
o
TRANCHANT et al., 2019). The SMOS data, are being established, aiming at strength- HYCOM+RODAS is now able to assimilate: they can support many other users asso-
however, are subject to large errors and are ening operational oceanography in Brazil (i) SSTfields; (ii) SSH or SSH anomaly (SSHA) ciated with the marine environment. For
not yet used operationally, despite their po- and abroad. data gridded or along the satellite track; example, the TAMAR Project, dedicated to
tential for better understanding of process- As a result of REMO, the first operation- and (iii) T/S vertical profile data from Argo the conservation of sea turtles, used REMO
es such as the interaction of the Amazon al short-term ocean numerical forecasting profilers, CTDs, XBTs, PIRATA (MIGNAC et products to investigate the migration and
River plume with the western tropical At- system (up to 5 days) with assimilation of al., 2015; SANTANA et al., 2020; TANAJU- dispersion of sea turtles in the South Atlan-
lantic circulation (VARONA et al., 2019). In oceanographic data in Latin America was RA et al., 2020). tic (MANSFIELD et al., 2017).
2019, GODAE OceanView was terminated implemented at CHM in August 2009. The The positive impact that data assimila- The analyses complement efforts in
and OceanPredict started. Today, 12 coun- first results were publicly offered in the tion can have on the prediction of thermo- the implantation and improvement of ob-
tries participate in OceanPredict: Australia, beginning of 2010. This achievement led haline structure and SST and SSH anomalies servational systems through the so-called
Brazil, Canada, South Korea, China, USA, REMO to be admitted to the GODAE Ocean- of the HYCOM+RODAS system over a 30- experiments with observational systems
France, Italy, India, Japan, and Norway, in View in late 2010. The prediction system day horizon at Metarea V was demonstrat- (OSEs - Observing System Experiments)
addition to the European Community. In with assimilation has been improved over ed in Carvalho et al. (2019). At the end of and experiments to simulate observational

438 BLUE ECONOMIY From Observation to Data Use 439