Articles— Article I is reproduced with the permission of Elsevier, B.V. Article II is repro-duced with the permission of the Taylor & Francis Group. Article III is reprorepro-duced under the Creative Commons BY-NC-ND 4.0 International Licence (http://creativecommons.org/
licenses/by-nc-nd/4.0/). Article IV is reproduced under the Creative Commons BY 4.0 Li-cence (http://creativecommons.org/licenses/by/4.0/).
Figures— Fig. 1.2: National Archives of Finland (Luotsi- ja majakkalaitoksen arkisto, Uaab:139); http://digi.narc.fi/digi/view.ka?kuid=24767256. No copyright restrictions apply to the use of this photo.
Data— Finngrundet wave buoy data was obtained from the SMHI open data service (http:
//opendata-download-ocobs.smhi.se/explore/), licensed under the Creative Commons CC-BY-2.5 licence. Argo data were collected and made freely available by the International Argo Program and the national programs that contribute to it (http://www.argo.ucsd.edu, http://argo.jcommops.org). The Argo program is part of the Global Ocean Observing Sys-tem. Parts of the CTD monitoring data were extracted from the ICES Dataset on Ocean Hydrography (The International Council for the Exploration of the Sea, Copenhagen, 2014, http://ices.dk). The satellite SST data in Article II is available under the Creative Commons BY 4.0 International license and is processed by Finnish Environment Institute (SYKE).
This study has been conducted using E.U. Copernicus Marine Service Information. The Gulf of Finland Year 2014 data providers include Estonian Marine Institute (EMI); Marine Systems Institute (MSI); Finnish Environment Institute (SYKE); Uusimaa and South-East Finland Centres for Economic Development, Transport and the Environment (UUDELY and KASELY); City of Helsinki Environment Centre (HELSINKI); and North-West Inter-regional Territorial Administration for Hydrometeorology and Environmental Monitoring (HYDROMET). The bathymetric data in Article IV was obtained from the Finnish Inventory Programme for the Underwater Marine Environment (VELMU) depth model (Finnish En-vironment Institute, http://www.syke.fi/en-US/Open_information/Spatial_datasets) and the Baltic Sea Bathymetry Database (Baltic Sea Hydrographic Commission, 2013) version 0.9.3, downloaded from http://data.bshc.pro on 25 Aug 2014. VELMU data is licensed under the Creative Commons BY 4.0 licence and Baltic Sea Bathymetry Database (BSBD) data is li-censed under the Creative Commons BY 3.0 licence. The CTD monitoring data in Article IV was provided by Finnish Environment Institute (SYKE) and Centres for Economic Devel-opment, Transport and the Environment (ELY), and is licensed under the Creative Commons BY 4.0 licence, see http://www.syke.fi/en-US/Open_information. To view a copy of these licenses, visit http://creativecommons.org/licenses or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.
I
Vertical temperature dynamics in the Northern Baltic Sea based on 3D modelling and data from shallow-water Argofloats
Antti Westerlund⁎, Laura Tuomi
Finnish Meteorological Institute, Marine Research, Erik Palménin aukio 1, P.O. Box 503, FI-00101 Helsinki, Finland
a b s t r a c t a r t i c l e i n f o
Article history:
Received 15 July 2015
Received in revised form 14 January 2016 Accepted 20 January 2016
Available online 28 January 2016
3D hydrodynamic models often produce errors in the depth of the mixed layer and the vertical density structure.
We used the 3D hydrodynamic model NEMO to investigate the effect of vertical turbulence parameterisations on seasonal temperature dynamics in the Bothnian Sea, Baltic Sea for the years 2012 and 2013. We used vertical pro-files from new shallow-water Argofloats, operational in the area since 2012, to validate our model. We found that NEMO was able to reproduce the general features of the seasonal temperature variations in the study area, when meteorological forcing was accurate. Thek-εandk-ωschemes were selected for a more detailed analysis. Both schemes showed clear differences, but neither proved superior. While sea surface temperature was better simu-lated with thek-ωscheme, thermocline depth was clearly better with thek-εscheme. We investigated the effect of wave-breaking on the mixing of the surface layer. The Craig and Banner parameterisation clearly improved the representation of thermocline depth. However, further tuning of the mixing parameterisations for the Baltic Sea is needed to better simulate the vertical temperature structure. We found the autonomous Baltic Sea Argofloats valuable for model validation and performance evaluation.
© 2016 Elsevier B.V. All rights reserved.
Keywords:
Vertical and horizontal density gradients are an essential part of ocean dynamics. To be considered good, a hydrodynamic model must be able to faithfully reproduce them. Several studies have shown that 3D hydrodynamic models often produce considerable errors in mixed layer depths and in vertical temperature structure that can be related to the parameterisations of vertical turbulence. These errors can be es-pecially pronounced in areas with complex hydrography, such as coast-al seas, where parameterisations that have been proven to work well in the World Ocean may fail. Although their performance can be improved by tuning the parameterisations (Meier, 2001), the models still struggle to produce the stratification with sufficient accuracy.
Accurate simulation of density stratification is important not only for the model performance per se, but also for several applications that use model results. For example, in biogeochemical models the accuracy of nutrient concentrations in the surface mixed layer strongly depends on the density stratification of the hydrodynamic model. As upward transport of nutrients is mainly physically driven (e.g.Reissmann et al., 2009), problems with density stratification can lead to incorrect estimates of nutrient concentrations in the mixed layer. This has a direct effect on the representation of the primary production in the euphotic zone. Furthermore, biogeochemical model parameters have been shown to exhibit high dependence on the chosen turbulence closure
(Burchard et al., 2006). Other relevant applications include the calcula-tion of sound speed in the sea, which is based on temperature and salin-ity profiles (e.g.Apel, 1987).
The Baltic Sea is a semi-enclosed brackish water basin with specific horizontal and vertical stratification conditions compared to those of the World Ocean. The density stratification in the Baltic Sea is mostly de-termined by salinity. The water body of the Baltic Sea has a permanent two-layer structure as a result of saline water inflow from the Danish Straitsfilling the deeps, and of voluminous river runoffs bringing fresh water to the surface. The halocline is usually at a depth of 40–80 m.
The seasonal thermocline starts to develop in late spring and reaches its maximum depth of 10–30 m typically in late August. In the autumn the thermocline vanishes due to convection caused by the cooling and wind-induced mixing. The location of thermocline and halocline at dif-ferent depths produces stratification conditions that are challenging for the present ocean models, as was shown e.g. byMyrberg et al. (2010) andTuomi et al. (2012), who suggested further effort to develop model-ling methods for vertical mixing with particular relevance for the Baltic Sea.
Our study area, the Bothnian Sea (Fig. 1), is a semi-enclosed basin in the Gulf of Bothnia, Baltic Sea. Gulf of Bothnia consists of the Archipelago Sea, the Åland Sea, the Bothnian Sea and the Bothnian Bay. The Gulf of Bothnia has sills and archipelagos in the south to the Baltic Proper.
The hydrography of the Gulf of Bothnia differs considerably from that of the other basins of the Baltic Sea. The salinity stratification in the Bothnian Sea depends on the water exchange between the Baltic Proper and the Bothnian Sea through the Åland Sea and the Archipelago Sea