She et al. 
Operational Oceanography and Earth System Science 
Frontiers In Earth Science | www.frontlersln.org 
8 
February 2020 | Volume 8 | Article 7 
has increased relatively more than oxygen consumption in the 
sediment. Subsequently, natural ventilation has become less 
effective representing a positive feedback for hypoxia (Meier 
etal., 2018b). 
Observations from field campaigns from the northern Baltic 
Sea suggested that the flow regimes are intermittent and that 
hydraulic control occurs in only about 55% of the cases, i.e., less 
frequently than anticipated (Green et al., 2006). Further, in wider 
gravitational flows, transverse Ekman circulation was identified 
to be an important process for the generation of mixing (Umlauf 
and Arneborg, 2009a,b). 
Recently, there has been increased research into the Baltic 
Sea coastal zone, particularly into upwelling, nutrient retention 
and the coastal filter capacity of nutrients (e.g., Edman et al., 
2018). Estimates suggest that the coastal filter of the entire 
Baltic Sea removes 16% of nitrogen and 53% of phosphorus 
inputs from land (Asmala et al., 2017). Simulated long-term 
nutrient retention was found to be associated with the physical 
characteristics of a water body, such as the surface area, depth 
and residence time of the water. 
Progress was also made in understanding the large-scale 
circulation, water mass transformations, and mixing processes 
in the Baltic Sea using high-resolution ocean circulation models 
that were running for many decades together with Eulerian 
concentration and age tracers (e.g., Meier, 2005, 2007). The 
model results illustrate possible pathways and ages of either 
inflowing saline water from the North Sea or freshwater 
originating from the various rivers. Freshwater is found to be 
subject to an efficient recirculation in the Baltic (e.g., Rodhe 
and Winsor, 2002). These simulations are complementary to an 
interesting tracer release experiment in the deep water of the 
central Gotland Basin showing a considerable increase in vertical 
mixing rates after the tracer reached the lateral boundaries of the 
basin (Holtermann and Umlauf, 2012; Holtermann et al., 2012). 
Hence, boundary mixing is perhaps the key process of basin-scale 
vertical mixing. For further details, the reader is referred to the 
review article by Omstedt et al. (2014) and the original literature 
cited therein. 
Climate and Environmental Observations 
and Reanalyses 
Nowadays, meteorological databases (both station data and high- 
resolution gridded datasets) are freely available with high quality 
to force ocean models on decadal and even centennial time 
scales. For instance, the regional reanalysis project Uncertainties 
in Ensembles of Regional ReAnalyses (UERRA, http://www. 
uerra.eu) delivers homogenous atmospheric surface fields for 
the period 1961 until today. In addition, oceanographic data 
became more easily accessible and new important measurement 
platforms, such as the MARNET stations (https://www.io- 
warnemuende.de/marnet-en.html), long-term moorings, e.g., 
in the Gotland Deep region, FerryBoxes, and satellites, have 
provided temporally and spatially better resolved observations. 
River runoff data are crucial for the understanding of the Baltic 
Sea dynamics and new catchment-wide high-resolution datasets 
based on process-based hydrological modeling calibrated to 
available station data are now available. However, a homogeneous 
hydrological dataset that covers the entire period from the 
1960s to the present day comparable to atmospheric reanalysis 
data is still missing. Further, available nutrient load and other 
environmental data are nowadays collected and stored in publicly 
available databases. 
A big step forward to understand climate variability in 
the Baltic Sea region was the development of historical 
reconstructions of atmospheric, hydrological and oceanic 
datasets since around 1850. With the help of Baltic Sea models, 
the impact of increasing nutrient loads and climate change on 
the marine ecosystem was detected and attributed to the various 
drivers of the system. We have now a better understanding of the 
natural variability in the Baltic Sea region and how large-scale 
atmospheric circulation affects the Baltic Sea climate variability 
(e.g., Borgel et al., 2018). During recent decades, changes in large- 
scale atmospheric circulation have caused a north-eastward shift 
in low-pressure tracks consistent with a more zonal circulation 
over the Baltic Sea basin (e.g., Trenberth et al., 2007). The decadal 
and multi-decadal regional variability of the past climate is partly 
explained by the North Atlantic Oscillation (NAO, mainly during 
winter) and the Atlantic Multi-decadal Oscillation (AMO). 
Despite the pronounced internal variability, trends were detected 
that could probably be attributed to anthropogenic climate 
change on centennial time scale (e.g., Kniebusch et al., 2019a,b). 
A highlight was the revision of the empirically derived barotropic 
saltwater inflow statistics for 1887 until present that shows 
no statistically significant trend but the same multi-decadal 
variability as in precipitation data (Mohrholz, 2018). Further, 
based upon model results it was concluded that stagnation 
periods such as the one between 1983 and 1992 are part of the 
natural variability of the system and occur once per 100 years on 
average (Schimanke and Meier, 2016). 
While atmospheric reanalysis data have long been used to 
force ocean models, long-term reanalyses for the ocean on multi- 
decadal time scales became only recently available including both 
physical and biogeochemical variables (e.g., Liu et al., 2017). 
Ocean reanalysis data play an important role for the development 
and evaluation of ocean models (Placke et al., 2018). 
Climate and Environmental Modeling 
Within BALTEX, the first coupled atmosphere-ice-ocean regional 
models were developed about 20 years ago (Gustafsson et al., 
1998; Hagedorn et al., 2000; Doscher et al., 2002; Schrum et al., 
2003). Nowadays several coupled models for the Baltic Sea— 
North Sea system are under development (e.g., Groger et al., 
2013; Tian et al., 2013, 2016; Van Pham et al., 2014; Wang 
et al., 2015; Ho-Hagemann et al., 2017). Regional climate models 
contributed to a better quantitative understanding of the energy 
and water cycles of the Baltic Sea basin. However, especially 
processes important for the regional water cycle are still not 
well-understood causing, inter alia, precipitation and runoff 
biases over the catchment area in long-term atmosphere climate 
simulations with considerable impact on the quality of ocean 
climate simulations (Meier et al., 2019). 
These models will be the future tools to investigate the 
dynamics of regional Earth systems. Within Baltic Earth a