She et al. 
Operational Oceanography and Earth System Science 
Frontiers In Earth Science | www.frontlersln.org 
6 
February 2020 | Volume 8 | Article 7 
ffi Clear Selection 
1 PARAMETERS < 
® PLATFORM TYPE v 
□ 
• 
ARGO/PROFILER 
□ 
О 
BATHY MESSAGES ON GTS 
□ 
® 
CTD PROFILES 
□ 
m 
DRIFTING BUOYS 
□ 
• 
FERRYBOX/SHIP 
□ 
GLIDERS 
□ 
© 
MARINE MAMMAL 
□ 
® 
MINI LOGGER 
□ 
® 
MOORING TIME SERIES 
0 
® 
PROFILING MOORING 
□ 
• 
RADAR 
□ 
® 
RIVER STATION 
□ 
• 
TESAC MESSAGES ON GTS 
□ 
• 
TIDE GAUGE 
□ 
• 
XBT OR XCTD PROFILES 
0 
<s> 
TIDE GAUGE GPS 
1? DEPTH 
m 
iti 
Trondelag 
Hedmark 
:Romsdol 
ry 
09fi fW 
me / 
^Norge 
*Y~ 
md 
bverige 
jomtlonds Vasternorrlandl’ 
Ion Ian + 
Dolarnos 
Ian 
iorÄ/s/fO 
Govleborgs ' ä 
ion [ / Pohjanl^itij : 
О • • 
f ( /V 
Suomi 
В Lonsi Suomi 
ItoSuomi 
Oslo 
yvormlonds 
i \ l0n 
. )srr ■ 
\ Cotalonds J 
г *л 
V. ••• 
Danmaÿo ^ 
j* 
Stc^ÿholnj 
тж 
EteloSuomi a 
H к 
• a ^Tallinn 
• • 
- 
»? 
Eesti 
r " 11a î •,,*! • * 
• * Gd ^ k 
jL^°' srein jfäUfkienäm^ 
- "{ \ ororpommerW^ 
tangen Hamburg , Bydgoszcz 
Vlgitvija 
Lietuva 
в Калининград ® 
Vilnius 
Гродно 
- \° 
. Псковская. 
I .область 
Витебская 
облость 
Беларус 
FIGURE 1 I Station distribution of operational observations In the Baltic Sea. 
Nielsen, 1999; Tuomi et al., 1999) based on the WAM model. 
Ocean, ice and oil drift forecast models BSH-Cmod, BSH-Dmod, 
HIROMB, and HELMI had been developed and operationalized 
in the early and mid-1990s (Haapala and Lepparanta, 1996; Dick 
et al., 2001; Wilhelmsson, 2002; Funkquist and Kleine, 2007). 
They are currently updated by more advanced coupled ocean- 
ice forecasting systems HBM (HIROMB-BOOS Model, Berg and 
Poulsen, 2012), NEMO-Nordic (Hordoir et al., 2018), and GETM 
(General Estuarine Transport Model, Burchard and Bolding, 
2002; Biichmann and Soderkvist, 2016). HBM is a dynamically 
two-way nested model with excellent hybrid parallel computing 
performance (Poulsen et al., 2014). NEMO is the European 
operational model with the largest user community. GETM 
has advantages in resolving the coastal-estuary continuum with 
specific advances in turbulence closure schemes and reduced 
diapycnal mixing due to the usage of vertically adaptive 
coordinates (Burchard et al., 2009). Biogeochemical models 
such as ERGOM (Neumann, 2000; Maar et al., 2011) and 
SCOBI (Swedish Coastal and Ocean Biogeochemical model, 
Eilola et al., 2009) have been developed in this century for 
setting up operational ecological service. The former has been 
used to provide basin-scale biogeochemical forecasts for CMEMS 
since 2009 (Tuomi et al., 2018) while the latter was used for 
producing biogeochemical reanalysis (Liu et al., 2017). The 
above operational model systems have been applied for basin- 
scale forecasts in 0.5-1 nautical mile (nm) resolution and 
a up to 60 m resolution for local scale forecasts (She and 
Murawski, 2018). Coupled model development, especially wave 
related coupling processes, e.g., ocean-wave, atmosphere-wave, 
and wave-ice interaction, is an on-going activity by BOOS 
partners. In the Phase II of CMEMS BAL MFC (2018-2021), 
a fully coupled ocean-wave-ice-biogeochemical model system 
NEMO-LIM-WAM-ERGOM together with PDAF (Parallel Data 
Assimilation Framework) assimilation is under development. 
Data Assimilation 
Significant data assimilation capacity has also been developed 
in the operational community, ranging from a simplified 
Kalman Filter for sea surface temperature (SST) assimilation 
in BSH-Cmod (Larsen et al., 2007), pre-operational 3DVAR 
(3D Variational method), and EnOI (Ensemble OI) for T/S 
assimilation and reanalysis production in HBM (Zhuang et al., 
2011; Fu et al., 2012) to the ensemble variational method 
and PDAF simplified Karman filter for physical-biogeochemical