The Oxford Research Encyclopedia of Climate Science will be available via subscription on April 26. Visit About to learn more, meet the editorial board, or recommend to your librarian.

Dismiss
Show Summary Details

Page of

date: 24 April 2018

# Future Climate Change in the Baltic Sea Region and Environmental Impacts

## Summary and Keywords

The warming of the global climate is expected to continue in the 21st century, although the magnitude of change depends on future anthropogenic greenhouse gas emissions and the sensitivity of climate to them. The regional characteristics and impacts of future climate change in the Baltic Sea countries have been explored since at least the 1990s. Later research has supported many findings from the early studies, but advances in understanding and improved modeling tools have made the picture gradually more comprehensive and more detailed. Nevertheless, many uncertainties still remain.

In the Baltic Sea region, warming is likely to exceed its global average, particularly in winter and in the northern parts of the area. The warming will be accompanied by a general increase in winter precipitation, but in summer, precipitation may either increase or decrease, with a larger chance of drying in the southern than in the northern parts of the region. Despite the increase in winter precipitation, the amount of snow is generally expected to decrease, as a smaller fraction of the precipitation falls as snow and midwinter snowmelt episodes become more common. Changes in windiness are very uncertain, although most projections suggest a slight increase in average wind speed over the Baltic Sea. Climatic extremes are also projected to change, but some of the changes will differ from the corresponding change in mean climate. For example, the lowest winter temperatures are expected to warm even more than the winter mean temperature, and short-term summer precipitation extremes are likely to become more severe, even in the areas where the mean summer precipitation does not increase.

The projected atmospheric changes will be accompanied by an increase in Baltic Sea water temperature, reduced ice cover, and, according to most studies, reduced salinity due to increased precipitation and river runoff. The seasonal cycle of runoff will be modified by changes in precipitation and earlier snowmelt. Global-scale sea level rise also will affect the Baltic Sea, but will be counteracted by glacial isostatic adjustment. According to most projections, in the northern parts of the Baltic Sea, the latter will still dominate, leading to a continued, although decelerated, decrease in relative sea level. The changes in the physical environment and climate will have a number of environmental impacts on, for example, atmospheric chemistry, freshwater and marine biogeochemistry, ecosystems, and coastal erosion. However, future environmental change in the region will be affected by several interrelated factors. Climate change is only one of them, and in many cases its effects may be exceeded by other anthropogenic changes.

# Introduction

During the last few decades, it has become increasingly clear that humans are affecting the global climate. The global mean temperature has increased by nearly 1°C since the late 19th century, and anthropogenic emissions of carbon dioxide (CO2) and other greenhouse gases have been identified as the main culprit (Bindoff et al., 2013). Much larger climate changes might occur during the 21st century and beyond, particularly if ongoing political efforts to reduce greenhouse gas emissions are not successful.

Since climate widely affects the natural environment and human activities, adaptation to changing climate will be needed. To help the adaptation, information on the possible characteristics of the future climate change and on the sensitivity of natural and human systems to this change is required. This article outlines the current knowledge on future climate change for the Baltic Sea and its drainage basin in northern Europe. The potential environmental impacts of climate change in the area are also discussed, although in less depth than climate change itself.

The Baltic Sea is a large, semi-enclosed brackish water estuary adjoining the northeastern Atlantic Ocean, with an area of 415,000 km2 with Kattegat included. The Baltic Sea drainage basin covers 1.7 million km2 and hosts 85 million people in 14 countries (HELCOM, 2015). Given this large population, it is unsurprising that climate change is not the only environmental issue in the area. In particular, the ecological state of the Baltic Sea has been severely degraded by water- and airborne nutrient loads and other pollution from the surrounding land areas. This far, the environmental effects of climate change have been in many cases secondary to these direct anthropogenic impacts. Similarly, the question of how the environment in the Baltic Sea region will change in the future is much wider than how it will be affected by climate change alone. Nevertheless, climate change already represents an additional stress factor for many natural systems that is likely to become more important in the future.

This article first discusses the historical background of climate change research for the Baltic Sea area, outlines the modeling tools and scenarios used for constructing projections of future climate change, and puts the climate changes projected for the Baltic Sea region in a global perspective. After this, climate change in the Baltic Sea basin is discussed in more detail, covering various aspects of the atmospheric climate, snow conditions, hydrology, and ocean climate in the Baltic Sea itself. Third, the potential environmental impacts of climate change are explored. For the environmental impacts, in particular, this text only scratches the surface of a very wide and complicated topic. Many of the issues introduced in this article have been discussed in more depth in the first (BACC Author Team, 2008) and the second (BACC II Author Team, 2015) climate change assessments for the Baltic Sea basin.

# Historical Context

The possibility that changes in atmospheric CO2 concentration could affect the global climate was already raised in the 19th century (Arrhenius, 1896). Measurements started in 1958 by Charles Keeling at Mauna Loa in Hawaii showed that the CO2 concentration was actually increasing, which is now convincingly ascribed to human activities, mainly burning fossil fuel (Ciais et al., 2013; Watson, Rodhe, Oeschger, & Siegenthaler, 1990). Finally, in 1995, comparison between climate model simulations and observed temperature changes allowed the Intergovernmental Panel on Climate Change (IPCC) to declare that “The balance of evidence suggests a discernible human influence on global climate” (IPCC, 1996, p. 4)—a conclusion that has been substantially strengthened by later research. These facts raise two important questions. First, how will climate change in the future? Second, how will the change affect humans and the natural environment?

Climate will not change in the same way in all parts of the world. Although more or less all areas are expected to become warmer, the magnitude of the warming will vary. For other variables, such as precipitation, even the sign of the change will differ between regions. Estimates on the regional distribution of future climate change have been available from three-dimensional global climate models (GCMs) since the 1980s (Stouffer, Manabe, & Bryan, 1989).

The first GCMs were still very coarse-grained and their results thus allowed only broad and tentative conclusions about the regional distribution of climate changes. Nevertheless, many of the conclusions have stood the challenge of time. For example, the models available for the first IPCC assessment consistently indicated a greater than global average warming at high northern latitudes in winter, and increases in precipitation at high latitudes throughout the year and at mid-latitudes in winter (Mitchell, Manabe, Meleshko, & Tokioka, 1990, p. 135). Since then, climate models have become much more sophisticated, with many more processes included and finer horizontal resolution. The number of models developed in different research institutions around the world has also greatly increased. Nonetheless, the models used in the IPCC fifth assessment still nearly unanimously agreed on the mentioned large-scale features of temperature and precipitation change (Collins et al., 2013, pp. 1059 and 1078).

Regional studies that specifically explored potential future climate change in areas close to the Baltic Sea region also began to appear in the 1990s. In the first multi-GCM study for northern Europe, Räisänen (1994) used four GCMs forced with a gradual doubling of atmospheric CO2. He found the average simulated winter warming in Finland to be 3 to 4°C, whereas the warming in summer was slightly over 2°C. Because these simulations were based on a highly idealized forcing scenario, their results are not directly applicable to the real world. However, similar idealized GCM experiments conducted in the early 2010s still largely reproduced these early results, although with a tendency toward slightly larger summer warming (Räisänen & Ylhäisi, 2015).

The realization that climate was changing and might change much more during the next hundred years soon stimulated a surge of research on the impacts of climate change. An early national example was the Finnish Research Programme on Climate Change (SILMU; Kämäri, 1997), which explored the potential effects of climate change on Finnish forests, peatlands, agriculture, inland waters, and the Baltic Sea. Guided by the GCM results available at that time, three idealized scenarios of climate change were formulated for the impact assessment (Carter, Posch, & Tuomenvirta, 1996). The scenarios assumed a linear 0.1 to 0.6°C per decade increase in the annual mean temperature and a 0.25 to 1.5% per decade increase in annual precipitation throughout the 21st century, with larger changes in winter than in summer in both of the variables. Another early example of impact research was the Climate Change and Energy Production project (Sælthun, 1992; Sælthun et al., 1998), which assessed the hydrological effects of climate change in the Nordic countries, with a special emphasis on hydropower projection, using temperature and precipitation change scenarios formulated by Jóhannesson, Jónsson, Källén, and Kaas (1995).

Nevertheless, the coarse resolution of GCMs (of the order of 500 km in the 1990s) was perceived as a serious limitation for regional climate change studies. At this resolution, geographic features, such as the Scandinavian mountains and the Baltic Sea, were at best marginally present in the GCMs. This motivated the development of higher-resolution regional climate models (RCMs). By the end of the 1990s, four countries in the Baltic Sea region—Sweden (Rummukainen et al., 2000), Norway (Bjørge & Haugen, 1998), Denmark (Christensen, Christensen, Machenhauer, & Botzet, 1998), and Germany (Jacob & Podzun, 1997)—had established their own regional climate modeling programs. Christensen et al. (2001) provided the first multi-RCM intercomparison of climate change simulations for the Scandinavian region. In the next decade, RCM simulations for the European area were conducted in a coordinated manner within the PRUDENCE (Christensen, Carter, & Rummukainen, 2007) and ENSEMBLES (van der Linden & Mitchell, 2009) projects. In the 2010s, an ensemble of very-high-resolution (12.5 km) RCM simulations for Europe was produced under the EURO-CORDEX initiative (Jacob et al., 2014).

Research on climate change and its impacts has been done in all Baltic Sea countries. To collect the pieces together, in 2008 an international team of authors compiled Assessment of Climate Change for the Baltic Sea Basin (BACC Author Team, 2008). Seven years later, an update was published (BACC II Author Team, 2015) to take advantage of the new knowledge gathered since the first assessment. The two BACC assessments are comprehensive, covering both past and potential future climate changes in the Baltic Sea region and the environmental impacts of the changes. In the second assessment, some of the potential socioeconomic impacts of climate change are also discussed.

# Models and Scenarios

It might be tempting to predict future climate change simply by extrapolating observed trends. However, such extrapolation would only make sense if past and future climate changes had exactly the same causes. In reality, anthropogenic climate change is the net effect of several factors—for example, changes in the concentrations of CO2 and other greenhouse gases, atmospheric aerosols, and land use—and the balance between them might differ between the past and the future. Moreover, climate exhibits substantial natural variability—temperature, precipitation, and other climate variables vary from year to year and from decade to decade, regardless of what humans are doing. Thus, climate changes, for example, during the past half-century represent a combination of anthropogenic change and natural variability, but it is often difficult to separate these contributions from each other. Similarly, future climate changes will result from a combination of anthropogenic change and natural variability, but it is not known whether the natural variations will oppose or reinforce the anthropogenic changes.

The main tools for constructing projections of future global and continental-scale climate change are global climate models (GCMs). The models are built on basic physical principles, such as conservation of mass, momentum, and energy, that govern the behavior of the atmosphere and the oceans. As input, a GCM needs information on “external” factors that it cannot predict, including, for example, either the anthropogenic emissions or concentrations of CO2 and other greenhouse gases. In addition, the initial conditions of the simulation (that is, the state of the atmosphere, oceans, sea ice, and land surface in the beginning of the model experiment) are needed. However, after the external factors and the initial conditions have been specified, the model simulation is self-contained. Thus, no information on observed climate change is used when simulating future climate change with GCMs.

GCMs simulate the evolution of atmospheric weather in a global three-dimensional network of grid points, with a time step of tens of minutes. Their output thus includes time series of temperature, winds, and other weather parameters all around the world, not only near the surface but also higher up in the atmosphere. Obviously, however, the detailed daily or even year-to-year evolution of weather cannot be predicted decades in advance. What the models attempt to predict is climate, that is, the long-term mean values and other statistical properties of weather. In such long-term statistics, the effects of unpredictable shorter-term variability are greatly reduced, although they still remain nonnegligible even in multidecade averages (Deser, Phillips, Bourdette, & Teng, 2012; Erikssson, Omstedt, Overland, Percival, & Mofjeld, 2007).

A major limitation to GCMs is their relatively coarse resolution, which is necessary to keep the need for computing resources manageable. The CMIP5 generation of models used in the 2010s (Taylor et al., 2012) typically had horizontal grid spacing of 100 to 250 km, whereas the earlier models were even coarser. This is a limitation in simulating regional climate and climate change, particularly as GCMs are only skillful in simulating features of climate that are several times larger in scale than their nominal grid spacing (Grotch & MacCracken, 1991).

Regional climate models (RCMs) have higher spatial resolution than GCMs (typically 10 to 50 km) but cover a limited geographic area, a few thousand kilometers across. They therefore need boundary conditions that specify the time-evolving atmospheric state at their horizontal boundaries. In climate change projection, the boundary conditions are derived from GCM simulations. Thus, climate changes in a RCM simulation depend on the driving GCM as well as the RCM itself (Déqué et al., 2012). Nevertheless, the higher resolution of RCMs compared with GCMs allows a better description of geographic features, such as mountain ranges and the land-sea distribution, that affect the regional climate. Due to their higher resolution, the RCMs also better resolve the details of atmospheric weather systems, which enhances the simulation of extremes, such as storms and heavy precipitation.

Another technique for constructing scenarios of future climate on small scales is statistical downscaling. In this case, observed statistical relationships between the large-scale atmospheric state and the local conditions are combined with either GCM or RCM simulations of larger-scale climate, to infer what the large-scale conditions would imply on the local scale. With statistical downscaling, scenarios can be constructed for truly local conditions, below the scale resolved by RCMs. On the other hand, statistical downscaling relies strongly on the assumption that the relationship between the large-scale and local conditions remains unchanged despite climate change (Benestad, 2016; Wibig et al., 2015).

A complication in both GCM- and RCM-based studies of future climate are model biases, that is, differences between the simulated and observed present-day climate. Even in state-of-the-art RCM simulations, the biases in temperature may reach several °C, and precipitation biases of several tens of percent are common (Wibig et al., 2015). In the ideal case that the biases remain similar in the simulation of future climate, they have no large effect on the simulated temperature and precipitation changes. Nevertheless, such biases generally preclude the direct use of RCM output in, for example, hydrological models, without some form of statistical adjustment (Graham, Andréasson, & Carlsson, 2007; Räty, Virta, Bosshard, & Connelly, 2017). Because there is no unique “correct” method for doing the adjustments, this intermediate step introduces additional uncertainties to the projection of (for example) river runoff and consequently Baltic Sea salinity.

In general, the word projection is preferred over prediction when referring to model simulations of future climate change. This is because the projections are conditional on the evolution of greenhouse gas emissions and other anthropogenic forcing. The evolution becomes naturally increasingly uncertain with an increasing time horizon. To cope with the uncertainty, emission or forcing scenarios are constructed using different but internally consistent assumptions. In the 21st century, two sets of scenarios have been widely used for climate projection: the SRES (Special Report on Emissions Scenarios; Nakićenović & Swart, 2000) and the RCP (Representative Concentration Pathways; van Vuuren et al., 2011) scenarios. These scenarios cover a wide range of CO2 emissions and concentrations (Figure 1), but nearly all of them indicate a monotonic increase in CO2 concentration at least until the end of this century.

Click to view larger

Figure 1. Total anthropogenic CO2 emissions (left) and CO2 concentrations for the SRES (top) and RCP (bottom) scenarios. The figure is based on the best-estimate values given in Appendix II of IPCC (2001) and Appendix II of IPCC (2013).

The only exception is RCP2.6, in which a deep decline in CO2 emissions leads to a slow decrease of the CO2 concentration in the second half of the century. In addition to CO2, other greenhouse gases, anthropogenic aerosols, and land-use changes are also included in the SRES and RCP scenarios.

# Global Perspective on Future Climate Change

The warming of the global climate is projected to continue in the 21st century, but the magnitude of the change is uncertain (Figure 2).

Click to view larger

Figure 2. Global mean temperature change under the four RCP scenarios, as simulated by the models used by Collins et al. (2013). Left: Multimodel mean change relative to the mean temperature in 1986 to 2005. The observed temperature time series up to the year 2015 is given as the mean of the HadCRUT4 (Morice, Kennedy, Rayner, & Jones, 2012), GISS (Hansen, Ruedy, Sato, & Lo, 2010), and NCDC (Smith, Reynolds, Peterson, & Lawrimore, 2008) estimates. Right: Temperature changes in the individual models from 1986–2005 to 2031–2050 and 2081–2100. Unit is °C.

The choice between different RCP (or SRES) scenarios has only a moderate effect in the next few decades, but becomes the dominating source of uncertainty by the end of the century (see also Hawkins & Sutton, 2009). However, the projections from different GCMs also differ from each other. For the ensemble of models in Figure 2, the changes in global mean temperature from the years 1986–2005 to 2081–2100 vary from 0 to 2°C under the lowest RCP scenario (RCP2.6) and from 2.5 to 5°C under the highest RCP scenario (RCP8.5).

Two factors contribute to the inter-GCM differences in climate change under the same forcing scenario (Collins et al., 2013). First, there are genuine differences in the response of the models to external forcing, such as increases in greenhouse gas concentrations, as well as subtle differences in the implementation of the scenarios. Second, there is unforced natural variability in the simulated climate. This stochastic variability is a very minor source of uncertainty in multidecade to centennial projections of the global mean temperature. However, it is more important for climate change on regional and local scales, particularly for variables like precipitation or wind speed (Hawkins & Sutton, 2011; Pryor & Barthelmie, 2010). Such unforced variability also occurs in the real world and sets an upper limit for the predictability of future climate. In some cases, this variability may be strong enough to dominate over the effects of anthropogenic climate change throughout the 21st century (Maraun, 2013).

The projected warming is generally larger over land than over sea and over high than over low latitudes (Collins et al., 2013). It is thus unsurprising that most GCMs project the annual mean warming in the Baltic Sea region to exceed the global average (Figure 3).

Click to view larger

Figure 3. Comparison between global (horizontal axis) and Baltic Sea drainage basin (vertical axis) area and annual mean temperature changes from 1986–2005 to 2081–2100. The same GCM simulations are included as in Figure 2 (33 for RCP2.6, 42 for RCP4.5, 26 for RCP6.0, and 42 for RCP8.5).

The warming in the Baltic Sea region tends to increase with increasing global mean warming, but it is not a one-to-one correlation. In some models, the regional warming is twice the global mean. However, there are also a few model simulations in which the Baltic Sea region warms less than the world on the average, or even becomes colder. When present in GCM simulations, cooling in the Baltic Sea region adjoins larger cooling over the northern North Atlantic (e.g., model FIO-ESM in Fig. 12.9 of Collins et al., 2013). It is not clear whether such a dramatic outcome is a real (although unlikely) possibility or results from a deficiency in some specific GCMs. Although many GCMs simulate some local cooling over parts of the northern North Atlantic, due to reduced northward heat transport by the ocean circulation, this cooling only rarely extends over the surrounding land areas (Collins et al., 2013, Fig. 12.9).

# Projected Climate Change in the Baltic Sea Region

## Temperature

Most GCMs project a very large warming over the Arctic Ocean in late fall and winter, due to increased ocean-to-atmosphere heat flux through reduced ice cover and other amplifying feedbacks (Collins et al., 2013). This strong warming extends in a muted form to the Baltic Sea region, particularly its northern parts (Figure 4a).

Click to view larger

Figure 4. Temperature change (Δ‎T) and precipitation change (Δ‎P) from 1986–2005 to 2081–2100 under the RCP4.5 scenario in the 42 GCMs used by Collins et al. (2013). (a)–(d): Maps of multimodel mean change. (e)–(f) Seasonal cycles of Baltic Sea drainage basin mean temperature change and precipitation change. The black lines show the multimodel mean, and the grey bars the mean ± 1 standard deviation (thus, this range covers approximately two thirds of the model projections). The Baltic Sea drainage basin is delineated by the thick purple line in (a)–(d). For reference, the multimodel global mean warming for these simulations is 1.8°C.

The warming in summer is weaker but nonetheless exceeds the global average in most GCMs (Figure 4c; Räisänen & Ylhäisi, 2015). Projections from RCMs agree with this large-scale pattern (Christensen, Kjellström, & Zorita, 2015). However, the intermodel variation of both the GCM and the RCM projections is substantial even for the same emissions scenario (Figure 4e; Fig. 11.3 in Christensen et al., 2015), although some warming occurs in nearly all of the projections.

The warming in winter will likely be accompanied by a decrease in temperature variability, on both daily and interannual time scales (Kjellström et al., 2007; Nikulin et al., 2011; Räisänen, 2002; Räisänen et al., 2004; Ylhäisi & Räisänen, 2014). This decrease is partly connected to the geographic distribution of the projected temperature changes (Holmes, Woollings, Hawkins, & de Vries, 2016). Larger warming over the Arctic Ocean and northern Eurasia than over the northern North Atlantic reduces the time mean temperature gradient, making temperatures less sensitive to variations of the atmospheric circulation. As a result, the lowest winter temperatures are projected to increase even more than the winter mean temperature. In summer, the variability might either slightly increase or remain at its current level. Both GCM and RCM simulations support increased temperature variability in summer in central Europe (Holmes et al., 2016; Kjellström et al., 2007; Nikulin et al., 2011; Räisänen et al., 2004; Ylhäisi & Räisänen, 2014), but the signal is far less clear further north in the Baltic Sea region. Even if the variability remains at its current level, periods of hot weather are expected to become more severe and more common following the increase in the summer mean temperature (Nikulin et al., 2011).

## Precipitation

On the large scale, precipitation is projected to increase in high northern latitudes. This is both due to increased northward moisture transport in the atmosphere (which is mainly caused by the larger moisture content of warmer air but is also affected by changes in atmospheric circulation) and increased moisture supply from local evaporation (Collins et al., 2013; Held & Soden, 2006). Conversely, precipitation is projected to decrease in many subtropical and lower mid-latitude areas, including the Mediterranean region. However, the transition between increasing and decreasing precipitation occurs further north in summer than in winter. For the Baltic Sea basin, this suggests a general increase in precipitation in winter (Figure 4b). In summer, precipitation may either increase or decrease, with a larger chance of drying in the southern than in the northern parts of the region (Figure 4d).

Projections of precipitation change are less robust than those of temperature change, varying widely between different GCMs and RCMs (Figure 4f; Fig. 11.5 in Christensen et al., 2015). This reflects, in part, the large natural variability of precipitation—in comparison with this variability, the projected anthropogenic precipitation changes are smaller than the changes in temperature (Hawkins & Sutton, 2011; Räisänen & Ruokolainen, 2006). Moreover, changes in atmospheric circulation may induce substantial variations in precipitation change on local scales, but this variability is smoothed out when averaging the projections of different models (Knutti, Furrer, Tebaldi, Cermak, & Meehl, 2010). Thus, much larger local increases or decreases in precipitation might occur than are suggested by Figure 4 (see Fig. 11.5 in Christensen et al., 2015).

Following the increase in saturation humidity with temperature, warming is generally accompanied by an increase in atmospheric water vapor (Held & Soden, 2006). Therefore, heavy short-term precipitation is likely to increase, even in many regions where the mean precipitation decreases. For the Baltic Sea region, most RCMs suggests an increase in extreme daily precipitation both in winter and in summer (Christensen et al., 2015, Fig. 11.6; Nikulin et al., 2011). This also holds for the southern parts of the basin, where the same models generally project a decrease in mean summer precipitation. On the other hand, model simulations suggest a decrease in the number of precipitation days in the southern and central parts of the Baltic Sea basin in summer (Lehtonen, Ruosteenoja, & Jylhä, 2014). This might increase the risk of long dry periods, particularly in the southern parts of the region (Orlowsky & Seneviratne, 2012).

## Wind

Future changes in wind speeds are more uncertain than the changes in temperature and precipitation. This holds for both the long-term mean and extremes of wind speed. The 13 RCMs used in the BACC II assessment show no clear preference for either increasing or decreasing wind speeds over land areas of the Baltic Sea basin during the 21st century, although local increases and decreases of ± 5% to 10% (or even more for extremes of wind speed) occur in some of the individual models (Christensen et al., 2015, Figs. 11.8–11.9). Ruosteenoja, Jylhä, and Kämäräinen (2016) analyzed the changes in mean wind speed in Finland in a large number of GCMs, obtaining largely similar results.

Changes in wind speed are highly dependent on changes in the large-scale atmospheric circulation. The upper row in Figure 5 shows an example of an RCM simulation with a very strong circulation change.

Click to view larger

Figure 5. Changes in annual mean sea-level pressure (left) and wind speed (right) from the years 1961–1990 to 2071–2100. The results are based on the SRES A2 emissions scenario and were produced using the same RCM (Rossby Centre regional Atmosphere-Ocean model; RCAO) using boundary data from two GCMs: ECHAM4/OPYC3 (top) and HadAM3H (bottom). Redrawn from Räisänen et al. (2004).

A pronounced decrease in sea-level pressure over the Arctic Ocean combined with an increase over central Europe indicate both an increase in the climatological north–south pressure gradient and a northward shift of cyclone activity. Both of these factors enhance wind speeds in northern Europe. By contrast, the simulation in the bottom row suggests only very small changes in the pressure distribution and wind speeds. Future changes in the atmospheric circulation depend on a multitude of partly compensating factors and are therefore difficult to predict (Shepherd, 2014).

In contrast with land areas, most RCM simulations suggest a slight increase in mean wind speed over the Baltic Sea (Christensen et al., 2015, Fig. 11.8; see also Figure 5). This is associated with reduced winter sea ice, which favors stronger near-surface wind speeds by weakening the stability of the atmospheric boundary layer. In many RCMs, a local maximum in wind speed change also occurs over the Baltic Sea in summer, due to a strong projected warming of the Baltic Sea surface. However, such stability effects are most important when the winds are relatively weak. The projected changes in the strongest winds do not differ much between the Baltic Sea and the surrounding land areas (Christensen et al., 2015, Fig. 11.9).

## Snow

Climate models project an increase in winter precipitation in the Baltic Sea basin, but in a warmer climate a smaller fraction of precipitation falls as snow and midwinter snowmelt episodes become more common. In general, a decrease in snow cover duration and snow depth is to be expected, but the decrease will be less dramatic in the colder (northern and eastern) than in the milder (southern and western) parts of the region (Christensen et al., 2015, Fig. 11.10; Räisänen & Eklund, 2012). This is because typical midwinter temperatures in the coldest parts of the region will remain well below zero even after a moderately large warming, whereas warming in milder areas will lead to a much larger increase in the frequency of above-zero winter temperatures.

Figure 6 illustrates the snow dynamics for three locations (Kiruna in Swedish Lapland, Joensuu in eastern Finland, and Riga in Latvia) based on RCM simulations from the ENSEMBLES project.

Click to view larger

Figure 6. Simulated 30-year mean seasonal cycles (August to June) of temperature (T), precipitation (PR), snowfall (PRSN), and water equivalent of the snow pack (SWE) at three locations in the Baltic Sea basin: Kiruna in Swedish Lapland, Joensuu in eastern Finland, and Riga in Latvia. Black = 1980–2010, blue = 2025–2055, red = 2069–2099. The figure is based on 12 RCM simulations from the ENSEMBLES project under the SRES A1B emissions scenario (Räisänen, 2015). Closed (open) circles indicate months in which all 12 (10 or 11) models agree on the sign of the change relative to 1980–2010.

The simulated temperatures increase at all three locations and so does (in most or all models) total winter precipitation. However, in Riga and Joensuu, snowfall decreases throughout the winter, although less in Joensuu than Riga. In Kiruna, midwinter snowfall actually increases, although the snowfall in autumn and spring months is reduced. The amount of snow, as indicated by the water equivalent of the snow pack, is reduced in both Riga and Joensuu much more drastically than could be expected from the relatively undramatic decreases in midwinter snowfall. This is partly due to a later start of the snowfall season, partly because of stronger midwinter snowmelt (Räisänen & Eklund, 2012). In Kiruna, the increase in midwinter snowfall nearly compensates for these two factors, leading to only a modest decrease in snow amount at the height of the snow season in March. Even in Kiruna, however, snow comes later in the fall and melts earlier in the spring.

Even in the future, snow conditions will vary substantially from year to year. Individual snow-rich winters are still likely to occur, although they will become gradually less common (Räisänen & Eklund, 2012).

The heaviest individual snowfall events typically occur when the temperature is close to, or just below, zero (O’ Gorman, 2014). At such temperatures, the air can hold more moisture than in colder conditions but is still cold enough for the precipitation to fall as snow. As such temperatures are expected to remain common even in the future, only small changes in daily extremes of snowfall are projected for the northern and eastern parts of the Baltic Sea basin (Räisänen, 2015). Further southwest, where snowfall as a whole becomes substantially less frequent, a slight decrease in snowfall extremes is likely, but the extremes are reduced much less in percentage terms than the total winter snowfall.

## Hydrology

Future changes in river flow will be affected by changes in three main meteorological factors: precipitation, evapotranspiration, and snow conditions. The total annual runoff that eventually enters the Baltic Sea is set by the difference between precipitation and evapotranspiration, but the snow pack acts a major seasonal storage of water that strongly regulates the seasonality of runoff generation. Snow naturally plays a larger role in the colder northern parts than in the milder southern parts of the region (Graham et al., 2008; Sonnenborg, 2015).

In general, evapotranspiration is projected to increase in a warmer climate if sufficient water is available, although there is substantial uncertainty about the magnitude of the change (Collins et al., 2013, Fig. 12.25; Graham et al., 2008). However, in the northern parts of the Baltic Sea basin, increases in precipitation are expected to overcome the increase in evaportranspiration, resulting in larger total annual river discharge. Further south, the projected precipitation increase is smaller and more uncertain, and most model simulations suggest reduced annual mean river flow (Collins et al., 2013, Fig. 12.24; Graham et al., 2008). Although it appears more likely that the increases in the north will outweigh the decreases in the south than vice versa, the change in the total annual river flow to the Baltic Sea is quite uncertain (Graham et al., 2008, Fig. 3.34).

The seasonal cycle of river runoff is projected to change (Graham et al., 2008; Sonnenborg, 2015). In areas that are currently characterized by snowmelt floods in the spring, the floods are likely to occur earlier and to decrease in magnitude because less snow accumulates during the shorter winter. Conversely, increases in precipitation together with reduced storage in snow act to increase the mean runoff and risk of flooding in winter. The runoff in late spring and summer is likely to decrease in most of the basin, due to the earlier snowmelt, increased evapotranspiration, and, possibly, particularly in the southern parts, reduced summer precipitation. However, in lake-rich river basins there is a considerable time lag between the runoff in the upper parts of the basin and the river flow that eventually reaches the Baltic Sea (Vehviläinen & Huttunen, 1997). Changes in the seasonality of runoff generation will have their strongest and most immediate effects in the upper sub-basins, proceeding down the lake chain in the course of weeks and months. In regulated watersheds, the seasonality is further modulated by changes in regulation practices that may be undertaken, in part, as adaptation to climate change (Veijalainen, Dubrovin, Marttunen, & Vehviläinen, 2010).

# Changes in Baltic Sea Ocean Climate

The Baltic Sea has several characteristics that make the impacts of climate change potentially different, and in some respects more difficult to anticipate, than is the case with the wider world ocean. First, it is a semi-enclosed basin receiving abundant river runoff from the surrounding land areas. Because of this, the water is far less saline than in the world ocean and the salinity is sensitive to changes in both the river runoff and saltwater intrusions through the Danish Straits. Second, the exchange of energy, water, and momentum between the atmosphere and the Baltic Sea is modulated by seasonal ice cover, whose extent is highly sensitive to winter climate. Third, ongoing land uplift counteracts large-scale sea-level rise, particularly in the northern parts of the Baltic Sea. Finally, the ecological impacts of climate change in the Baltic Sea occur against a background of severe eutrophication caused by nutrient input from the surrounding land areas.

Five aspects of Baltic Sea ocean climate change are germane: changes in ice cover, water temperature, salinity, sea level, and storm surges. Ice cover and water temperature affect the climate in the land areas surrounding the Baltic Sea, particularly in the relatively densely populated coastal regions. Water temperature and salinity together determine the density stratification of the Baltic Sea water mass, and thereby modulate the circulation of water and nutrients. Salinity is also biologically important for a wide range of organisms dwelling in the Baltic Sea. Changes in mean sea level and the height of storm surges affect coastal erosion and need to be considered when assessing the needs and options for coastal management.

The Baltic Sea is affected by the atmospheric climate change that occurs over the Baltic Sea itself. It is also affected by changes in river discharge from the surrounding land areas and by the sea-level rise in the world ocean. A range of time scales are involved in the response of the Baltic Sea to climate change, with water temperature and ice cover reacting much faster to atmospheric temperature change than salinity responds to changes in freshwater input (Omstedt & Hansson, 2006). Key tools for assessing the changes in the Baltic Sea are three-dimensional regional ocean model simulations, which have been available since the turn of the 21st century (Graham et al., 2008, Section 3.8; Meier, 2015). Projections for the changes in ice cover and sea level have also been derived by many other methods.

## Ice Cover

Ice conditions in the Baltic sea are primarily determined by the mean winter temperature (Jylhä, Fronzek, Tuomenvirta, Carter, & Ruosteenoja, 2008; Luomaranta et al., 2014; Meier, Döscher, & Halkka, 2004; Omstedt & Hansson, 2006; Tinz, 1996). In a warmer climate, the ice cover will therefore become thinner, less extensive, and less long-lived.

The most commonly used measure of the Baltic Sea ice winter severity is the annual maximum ice extent (MIB). Several studies have derived empirical relationships between MIB and winter temperature using observations (Jylhä et al., 2008; Luomaranta et al., 2014; Omstedt & Hansson, 2006; Tinz, 1996). For example, Luomaranta et al. (2014) found the exponential relation

$Display mathematics$
(1)

where T is the mean temperature of the November–March season averaged over the Baltic Sea coast, A = (90.2 ± 4.2) × 103 km2 and B = (0.253 ± 0.015) (°C)−1. This implies a 22% decrease in ice extent for each 1°C of winter warming. On the other hand, totally ice-free winters are hard to reach, since some ice tends to form in the north of the Gulf of Bothnia and in the eastern extreme of the Gulf of Finland even during relatively short cold periods. Regional ocean model simulations for the Baltic Sea suggest that ice-free winters would remain very rare even under the strong greenhouse gas forcing that characterizes the SRES A2 scenario in the late 21st century (Meier, 2006). Nevertheless, the projected warming indicates a large increase in the frequency of mild ice winters already in the first half of the 21st century, together with a virtual elimination of ice winters that would have been classified as severe in the 20th century (Luomaranta et al., 2014).

## Water Temperature

Following the warming of the atmosphere, the water in the Baltic Sea will become warmer. However, the magnitude of this change is modulated by changes in ice conditions and vertical mixing. Vertical mixing itself is affected by changes in temperature and salinity, which together determine the density of seawater. However, in contrast with more saline ocean water, the brackish water in the Baltic Sea reaches its maximum density when the temperature is a few degrees above the freezing point. These interlinked and nonlinear processes allow the warming of the surface water to differ from the warming of the air above, both in its annual mean and its seasonal cycle. The vertical gradient of water temperature is also projected to change, with larger warming near the surface than at greater depth (Meier, 2015).

As a specific example, regional Baltic Sea model simulations suggest a pronounced maximum of sea surface warming in the Bothnian Bay and Bothnian Sea during summer (Meier et al., 2012). This results from the nonlinear relationship between temperature and water density (Hordoir & Meier, 2011). After the ice melt in spring, seasonal warming first increases the density of the surface water. This triggers thermal convection, which mixes the heat down in the water column. Therefore, the surface water only warms up slowly. However, once the temperature in the surface layer exceeds the point of maximum density, further warming suppresses vertical mixing. After this, the warming at the surface accelerates, because the heat input from the atmosphere is distributed in a much thinner water layer. In a warmer climate, the point of maximum density is reached earlier in spring. Consequently, a larger fraction of the spring- and summertime heat input stays near the surface. The net result is a larger summertime temperature increase in the surface water than either in the air above or in the water at greater depth.

## Salinity

Most climate change projections suggest an increase in total river runoff to the Baltic Sea. The local freshwater input given by the difference of precipitation and evaporation over the Baltic Sea might also increase to some extent (Rutgresson, Omstedt, & Räisänen, 2002). This increased freshwater supply would reduce the salinity of the Baltic Sea water. The decrease might be substantial, although its magnitude is rather uncertain. For example, in the ensemble of ocean model simulations of Meier et al. (2012), the volume-averaged Baltic Sea mean salinity decreases from 8 to nearly 6 g kg−1 during the 21st century. The pattern of the projected change roughly mirrors the current salinity distribution. Thus, the largest decrease occurs in the Danish straits region, where the present-day salinity is largest, and the smallest decrease in the eastern and northern parts of the Baltic Sea, particularly the Bothnian Bay, where the water is less saline.

## Sea-Level Change

The sea level in the world ocean is projected to rise due to loss of land-based ice masses and thermal expansion of ocean water (Church et al., 2013). On the other hand, the loss of the Fennoscandian ice sheet in the end of the latest glaciation is still causing ongoing land uplift due to glacial isostatic adjustment (GIA). The land uplift rate varies from near zero at the Schleswig-Holstein coast in the southwest to 10 mm yr−1 in the Bothnian Bay (Hill, Davis, Tamisiea, & Lidberg, 2010; Johansson, Pellikka, Kahma, & Ruosteenoja, 2014). Therefore, the relative sea-level rise will be largest along the southwestern and southern coasts of the Baltic Sea. Around the Gulf of Bothnia, the relative sea level is still projected to decrease during this century, except for the most extreme scenarios. However, the decrease will be slower than previously.

The rate of future global sea-level rise is not well known. In the IPCC fifth assessment report, the projections for the change from the 20-year period 1986–2005 to the year 2100 range from 28 to 61 cm for the RCP2.6 scenario to 52 to 98 cm for RCP8.5 (Church et al., 2013). However, some semi-empirical models, built on the connection between observed sea level and temperature changes, have given higher estimates of global sea-level rise, in extreme cases up to 200 cm by the year 2100 (Table 4 of Johansson et al., 2014). The largest uncertainty concerns the marine-based sectors of the Antarctic ice sheet, whose response to climate change is very difficult to model and might turn out to be larger than was expected in the IPCC fifth assessment. A study by DeConto and Pollard (2016) suggested that, under the high-end RCP8.5 scenario, the loss of Antarctic ice alone could raise the global sea level by up to 1 m by the year 2100, and up to 15 m by the year 2500.

Sea-level rise will not be uniformly distributed around the world, even when local land uplift or subsidence is disregarded (Church et al., 2013). The change in the Baltic Sea region is likely to be smaller than the global average (Grinsted, 2015; Johansson et al., 2014). In particular, ice melt from the Greenland ice sheet and smaller northern hemisphere high-latitude glaciers would only weakly affect the sea level in the Baltic Sea. Due to changes in the Earth’s gravity field and other geophysical effects, the sea-level rise associated with the loss of land ice is small (or even locally negative) in nearby regions, but slightly larger than the global average in the opposite hemisphere (Bamber & Riva, 2010; Slangen, Katsman, van de Wal, Vermeersen, & Riva, 2012).

As the best estimate, the absolute sea-level rise in the Baltic Sea (before accounting for land movement) would be about 80% of the global mean. Grinsted (2015) combines a scenario with a midrange 70 cm global sea-level change during the 21st century with the simultaneous land uplift, arriving at a relative sea-level rise of 60 cm near Hamburg, but a sea-level fall of 35 cm in the Bothnian Bay. Excluded from this estimate are local effects from changes in winds and sea ice. These local effects might act to increase the sea level in the northern parts of the Baltic Sea, by up to 20 cm in the northernmost Bothnian Bay in spring (Meier, 2015, Fig. 13.7). However, the large uncertainty in the future rate of global sea-level rise makes the estimates of local sea-level change also quite uncertain.

## Storm Surges

The impacts of sea-level change depend on both the mean change and changes in variability. Of particular importance are the high extremes caused by wind storms, known as storm surges. Future changes in both mean and extreme wind speeds are still rather uncertain, and the same applies to storm surges. Some model simulations suggest an increase in extreme storm surges, whereas others do not (Meier, 2006). However, a recent study, in which a high-resolution hydrodynamic model was driven by output from eight GCMs, supports an increase in storm surges as a consensus projection for the whole Baltic Sea in all seasons (Vousdoukas, Voukouvalas, Annunziato, Giardino, & Feyen, 2016). On the other hand, the magnitude of the projected increase (of the order of 5%, or 10 cm for long-term extremes) is relatively modest. Therefore, the absolute change in sea-level extremes in the future will probably be mainly determined by the change in the mean sea level, particularly in the southern Baltic Sea, where the largest increase in mean sea level is projected (Gräwe & Burchard, 2012).

The distribution of storm-surge heights has a heavy tail. This means that, even in the absence of climate change, surges might in rare cases reach much higher than even 100-year-long observational records would suggest. For example, in Travemünde at the German Baltic Sea coast, a surge in 1872 exceeded the normal water level by 3.2 m (Jensen & Müller-Navarra, 2008). This is about 1.0 m higher than any other surge in the time series that goes back to the 1820s.

# Environmental Impacts of Climate Change

Changes in climate affect the environment in several ways. Some of the effects are direct (for example, the effects of temperature and precipitation on vegetation survival and productivity), others indirect (for example, changes in marine ecosystems that follow when changes in nutrient input due to changed river runoff alter the relative competitiveness of different species). In addition to climate change, however, the environment in the Baltic Sea region is subject to nutrient pollution and many other anthropogenic pressures. Moreover, some changes would occur even in the absence of the human impact and climate change—for example, due to the continuing land uplift. Due to this complexity, it has been in most cases difficult to pinpoint the climate change contribution to the environmental changes observed this far (BACC II Author Team, 2015). Similarly, the environmental changes that will occur in the future will result from several interrelated factors, of which climate change is but one.

The potential environmental impacts of climate change in the Baltic Sea region include issues related to atmospheric chemistry, terrestrial and freshwater ecosystems, terrestrial carbon storage, Baltic Sea biogeochemisty, ecosystems in the Baltic Sea, and coastal erosion and coastline changes. All of these are discussed in more depth in the first (BACC Author Team, 2008) and the second (BACC II Author Team, 2015) BACC assessments.

## Atmospheric Chemistry

Major air pollutants in the Baltic Sea region include acidifying components, such as sulfur and nitrogen oxides and ammonia (NH3), together with ozone (Simpson et al., 2015). Large changes in the severity of air pollution and thus its impacts have taken place during the past century. The emissions increased dramatically during the 20th century and particularly after World War II. Beginning in the 1980s, however, control measures have reduced the emissions of, in particular, sulfur compounds.

Future changes in the level of air pollution are also expected to be primarily determined by policy- and technology-driven changes in emissions. Potential climate change effects are secondary, but some of them have been identified. For example, climate change may slightly increase nitrogen deposition within the Baltic Sea drainage basin (Langner, Andersson, & Engardt, 2009). Furthermore, higher temperatures may increase ammonia emissions from land sources owing to increased evaporation (Skjøth & Geels, 2013; Sutton et al., 2013). Finally, reduced ice cover might lead to increased shipping and thus increased shipping emissions of many pollutants over the Arctic Ocean, which would primarily affect the northern parts of the Baltic Sea basin (Tuovinen, Hakola, Karlsson, & Simpson, 2013).

## Terrestrial and Freshwater Ecosystems

Ecosystems on land, in inland lakes and rivers, and in the Baltic Sea are sensitive to climate (Dippner et al., 2008; Niemelä et al., 2015; Smith et al., 2008; Viitasalo et al., 2015). However, they are also affected by anthropogenic pressures, such as loss of natural habitat, increases in carbon dioxide and ozone concentrations, toxic chemicals, and increased nutrient input via the atmosphere and from agriculture and wastewater. Furthermore, the direct physical effects of climate change are compounded by food-web interactions and competition between species. Therefore, the effects of climate change on ecosystems are in many cases difficult to predict and to separate from other factors.

Although not the only facet of climate change, the greenhouse-gas-induced warming of climate is the “largest” of the expected changes—that is, more clearly discernible from natural variability than, for example, changes in precipitation (Hawkins & Sutton, 2011). It is therefore useful to highlight the effects of temperature change with a few examples. Higher temperatures are projected to lengthen the growing season for terrestrial plants, and to promote northward and upward shifts in the boundaries of species occurrence (Smith et al., 2008). The latter will mean the advance of species in colder areas where they have not thrived before, sometimes at the cost of the original species in that area, but in some cases also retraction near the warm limit of the range. Examples of the potential consequences are displacement of tundra plants by grasses and trees, and an increased share of broadleaf trees in boreal forests. Changes such as increases in treeline altitude and increased plant species diversity at mountain tops are already observable (Kullman, 2004, 2010).

Continuing with forests, increased temperatures are expected to increase the growth of coniferous trees in the northern parts of the Baltic Sea basin, but may reduce it in the south, where the growth tends to be limited by water availability (Lindner et al., 2010). However, the earlier dehardening associated with milder winters might make the trees more vulnerable to temperature backlashes in the spring in some areas (Jönsson, Linderson, Stjernquist, Schlyter, & Bärring, 2004). Another risk associated with warming is a northward expansion of pests and pathogens (Lindner et al., 2010). Finally, the increase in temperatures may make trees more prone to windthrow by delaying the freezing of the soil in the autumn (Peltola, Kellomäki, & Väisänen, 1999). On the other hand, the increase in CO2 concentration is thought to enhance the positive effects and reduce the negative effects of climate change on forests and plants in general (Lindner et al., 2010).

The projected atmospheric warming acts to increase the water temperature in inland lakes and shorten the ice season. Due to the earlier ice break-up, the seasonal warming in spring will start earlier and the stratified summer period (when the warm surface water is less dense than the colder water below) will lengthen. The earlier ice melt will also increase the amount of sunlight available for phytoplankton in early spring. The hydrological regime in lakes will also change (see the section “Hydrology”), with potentially major changes in the seasonality of water and nutrient input into lakes.

The warming of water will favor warm-water species at the expense of cold-water species. Together with the increased stratification, it could also enhance eutrophication. On the other hand, a shortening of the ice season reduces the risk of anoxia during winter and spring. On the whole, the effects of climate change on lake ecosystems are complex and will depend on the individual characteristics of lakes (Smith et al., 2008).

## Terrestrial Carbon Storage

Most studies point toward increased terrestrial carbon storage in the Baltic Sea basin in the future. However, the actual outcome is uncertain due to a number of compensating processes (Niemelä et al., 2015; Smith et al., 2008). Warming and increased CO2 concentration both favor higher primary production, particularly in the northern parts of the area where water availability is mostly not a limiting factor. However, higher temperatures also increase respiration and accelerate the decomposition of organic matter in the soil. Another question mark is the interaction between the carbon and nitrogen cycles (Churkina et al., 2010; de Vries & Posch, 2011). Anthropogenic nitrogen deposition has probably played a major role in enhancing ecosystem carbon sequestration in Europe. Anticipated decreases in nitrogen deposition due to more stringent emission control may thus act to reduce the carbon sequestration in the future. Besides these factors, changes in land use and forest management may also have a large impact on ecosystem carbon storage (Zaehle et al., 2007).

## Baltic Sea Biogeochemistry

The main limiting nutrients for primary production in the Baltic Sea are nitrogen and phosphorus (Smith et al., 2008). A fraction of their input to the Baltic Sea occurs as deposition from the atmosphere, but over three quarters of the nitrogen input and more than 90% of the phosphorus input is waterborne, entering the Baltic Sea via rivers or directly from point sources at the coastline (HELCOM, 2015). The input of nutrients is largely dependent on human activities, particularly agriculture, but also forestry and releases from municipal wastewater plants. After increasing dramatically for most of the 20th century, the nutrient input has started to decrease. Nevertheless, socioeconomic factors, such as increasing meat consumption, might make it harder to achieve further decreases in the future (Hägg, Humborg, Morth, Medina, & Wulff, 2010).

Nutrient input to the Baltic Sea is also modulated by variations in meteorological and hydrological conditions (HELCOM, 2011; Humborg et al., 2015). Both temperature and precipitation affect the fraction of released nutrients that reaches the Baltic Sea. Increases in precipitation, as projected at least for the northern parts of the Baltic Sea basin, tend to increase nutrient transport by increasing river runoff. On the other hand, retention processes that remove nutrients from water before it reaches the Baltic Sea may become more efficient with increasing temperature (Hong et al., 2012; Veraart, de Klein, & Scheffer, 2011). The seasonality of the nutrient input is also likely to change, because the projected changes in precipitation and snow conditions imply increased river runoff in winter but reduced runoff during late spring and summer (see the section “Hydrology”).

The increase in nitrogen and phosphorus input has increased the biological production in the Baltic Sea by a factor of 2 to 4 (Schneider, Eilola, Lukkari, Muller-Karulis, & Neumann, 2015). This eutrophication has dire ecological consequences (HELCOM, 2009; Schneider et al., 2015). It promotes toxic algal blooms and the development of near-dead deepwater areas due to oxygen depletion and production of hydrogen sulfide (H2S). Another potentially important environmental issue in the Baltic Sea is the acidification of seawater induced by increasing atmospheric CO2 concentration (Omstedt et al., 2012).

How will the eutrophication-related problems, and particularly the oxygen depletion, change in the future? If anthropogenic nutrient input is reduced following the current Baltic Sea Action Plan (HELCOM, 2007), the situation is projected to improve, although quite slowly. However, with constant or increasing nutrient input, even larger areas with oxygen depletion would develop (Meier et al., 2011; Schneider et al., 2015). This is partly due to the internal cycling of the nutrients that have accumulated in the Baltic Sea and its sediments during the past century, but projected climate changes also play a role. First, the solubility of oxygen in the Baltic Sea water decreases with increasing temperature. Second, the projected warming accelerates the recycling of organic matter in the Baltic Sea. Third, increases in river runoff may increase the transport of nutrients from land to the Baltic Sea. However, the net effect of runoff changes is uncertain because many projections suggest a decrease in runoff from the southern parts of the Baltic Sea basin (Collins et al., 2013, Fig. 12.24; Graham et al., 2008).

Climate change might also modify the biogeochemical conditions in the Baltic Sea by altering the frequency and magnitude of saltwater inflows from the North Sea. Some studies have suggested an increase in such inflows in the future, but the evidence is inconclusive (Meier, Höglund, Eilola, & Almroth-Rosell, 2017; Schimanke, Dieterich, & Meier, 2014). The consequences of the inflows are also complicated. Contrary to intuition, stronger and more frequent saltwater inflows might lead to an expansion of dead bottom areas and increased risk for toxic Cyanobacteria blooms, at least in some parts of the Baltic Sea (Meier et al., 2017).

Oceans currently absorb a quarter of the anthropogenic CO2 emissions (Ciais et al., 2013, Table 6.3). The resulting acidification alters the seawater chemistry and may have serious ecological consequences. During the past two centuries, the pH of the Baltic Sea surface water is estimated to have decreased by about 0.15. A further decrease in pH is expected in the 21st century. The magnitude of the decrease depends primarily on the magnitude of anthropogenic CO2 emissions, with a decrease of 0.4 pH units possible under a high-end emission scenario (Omstedt et al., 2012).

## Marine Ecosystems

Marine ecosystems in the Baltic Sea are strongly affected by eutrophication, toxic chemicals, and other anthropogenic pressures, such as overfishing. However, the ecosystems are also sensitive to climate. Climate change directly affects the metabolism, growth, survival, and productivity of individual organisms, but it also indirectly influences the structure of communities, for example by shaping the temporal and spatial match or mismatch of interacting species (Viitasalo et al., 2015). These indirect effects are particularly difficult to predict, especially as climate change is compounded by many other anthropogenic pressures.

Climate change is expected to affect the Baltic Sea ecosystems at least via increases in water temperature, reduced sea ice, and changes in salinity. Furthermore, changes in temperature and salinity together regulate the density of seawater, thus affecting its vertical mixing (Viitasalo et al., 2015). Some of the effects are relatively straightforward. For example, a decrease in sea ice extent and duration is a serious threat for ice-dwelling species, such as the Baltic ringed seal, which may lose much of its current winter habitat (Meier, Döscher, & Halkka, 2004). Other effects are more complicated but no less important. In particular, increases in water temperature together with reduced ice cover may lead to major changes in the seasonal succession and species composition of the phytoplankton community (Sommer & Lengfellner, 2008; Viitasalo et al., 2015). One consequence of this is an increased risk of Cyanobacteria blooms. On the other hand, a decrease in salinity might reduce phytoplankton productivity, at least in the Gulf of Bothnia. The decrease in salinity together with poor oxygen conditions in the deep basins would also disfavor the main Baltic Sea piscivore cod, with many cascading effects to lower levels in the food web (Viitasalo et al., 2015).

It is still unknown how the projected CO2-induced acidification will affect the Baltic Sea ecosystems. Although many key species in the Baltic Sea food web appear to be physiologically tolerant of the pH decreases expected in the coming century, the acidification may nevertheless change the relative competitiveness of different species and in this way restructure the ecosystems (Havenhand, 2012).

## Coastal Erosion and Coastline Changes

The Baltic Sea coastlines are affected by a wide range of geological processes, human activities, and climatic factors. The main climatic factors are wind, waves, storm surges, ice jams, and flooding (Łabuz, 2015). The response of the coastline to climate change depends on both the nature of the climate change and the characteristics of the coastline. In general, the submergent and soft coast relief of the southern Baltic Sea area is under most threat of retreat, particularly as the projected sea-level rise accentuates its vulnerability to storm surges and wave erosion. The bedrock-dominated coasts of Finland and northern Sweden are less vulnerable, particularly as the local land uplift counteracts the absolute sea-level rise.

Apart from the effects of sea-level rise, shortening of the ice season in a warmer climate will also increase the exposure of shorelines to wave and wind erosion. On the other hand, the erosion caused by the ice pack itself will be reduced. Another potentially important but uncertain factor is changes in the frequency and strength of wind storms. The direction of the wind during the storms is also important, because the height of the waves and the storm surge depend on the fetch of the wind over the open Baltic Sea.

# Conclusion

The main drivers and interactions that are expected to shape climate change and environmental change in the Baltic Sea region in the 21st century are summarized in Figure 7.

Click to view larger

Figure 7. A schematic overview of the main drivers and interactions that are expected to affect the climate and the environment in the Baltic Sea region in the 21st century. The arrow from “Drivers of climate change” to “Regional climate change” refers to local climate forcings, such as land-use change. The arrows from “Large-scale climate change” and “Glacial isostatic adjustment” to “Physical changes in the Baltic Sea” refer to sea-level change.

The ultimate drivers of climate change are divided into three categories—increase in atmospheric greenhouse gases, other anthropogenic forcing, and natural variability. Similarly, three aspects of large-scale climate change are identified that will affect the climate change in the Baltic Sea region—global warming, changes in atmospheric circulation, and changes in ocean circulation.

When considering what can and what cannot be confidently said about future climate change in the Baltic Sea region, it is essential to understand that

1. 1. Of the three drivers in the top row of Figure 7, only the increase in greenhouse gas concentrations is both reasonably well understood and predictable. Other types of anthropogenic forcing, such as changes in aerosol emissions or land use, may either amplify or counteract the greenhouse-gas-induced changes, but their net effect is difficult to predict. This is partly due to limitations in current scientific understanding, and partly due to the short atmospheric lifetime of aerosols that allows for potentially fast changes in aerosol concentrations. Both forced (solar variability and volcanic eruptions) and unforced (internal dynamics of the climate system) natural climate variability are generally considered to have only limited predictability.

2. 2. The best understood effect from increasing greenhouse gas concentrations is a thermodynamically governed widespread increase in atmospheric temperatures and moisture content, commonly called “global warming.” Projections for changes in atmospheric and oceanic circulation are more uncertain, both because of the complicated dynamics involved and because anthropogenic circulation changes are superimposed onto strong natural variability.

3. 3. The anthropogenic greenhouse gas forcing and the resulting global warming are expected to increase gradually with time.

Consequently, the most certain (or least uncertain) aspects of regional climate change in the Baltic Sea basin are those that are most strongly affected by the large-scale global warming. These include increases in atmospheric temperature and Baltic Sea water temperature; decreases in ice cover in the Baltic Sea, lakes, and rivers; and increasing sea level at the southern coasts of the Baltic Sea, where the large-scale sea-level rise is not counteracted by glacial isostatic adjustment. These changes can be considered nearly certain, at least after the first few decades when they might still be masked by natural variability. They might be reversed only in the unexpected case that the large-scale warming would be regionally cancelled by a dramatic decrease in the North Atlantic thermohaline circulation.

Shortening of the snow season and decrease in snow amount are both driven by increased temperature but are likely to be counteracted by increased winter precipitation. Consequently, changes in snow conditions are subject to somewhat larger uncertainty than the warming. The increase in winter precipitation is also at least partly driven by the large-scale warming, which tends to increase the atmospheric moisture content and hence enables larger moisture transport from the south and west. Similarly, the increase in atmospheric humidity favors stronger precipitation extremes. However, precipitation is highly sensitive to atmospheric circulation and its changes are therefore not as strongly governed by atmospheric thermodynamics as the temperature change. It is mainly for this reason that climate change projections are more uncertain for precipitation than for temperature. Finally, changes in wind climate are to a large extent dictated by the atmospheric circulation and are therefore among the most uncertain aspects of climate change.

These considerations have clear implications for the environmental impacts of climate change. The impacts that are most strongly related to warming are more predictable than those related to precipitation and, in particular, wind changes. Nevertheless, for the environmental changes, climate change is only one part of the story. In some cases, climate change may become a dominant driver of environmental change; in others, its effects might remain small compared with the direct human influence. Some drivers of climate change, such as changes in land use and in CO2, ozone, and aerosol concentrations, also have direct environmental effects.

How can the uncertainty in the climate change projections for the Baltic Sea region be reduced? More detailed modeling and deeper understanding of the regional atmospheric, oceanic, and hydrological processes will undoubtedly be helpful. However, a large part of the uncertainty originates from outside the Baltic Sea region. This uncertainty can only be reduced by an improved understanding and modeling of the processes that control the magnitude of the global-scale warming, changes in atmospheric circulation and the North Atlantic ocean circulation. A lot of hard work is still needed with all of these issues. Finally, part of the uncertainty is essentially irreducible. The current understanding suggests that natural climate variability is largely unpredictable in the long run, although the ocean circulation might give it some predictability for the first decade or so (Koenigk, König Beatty, Caian, Döscher, & Wyser, 2012). Similarly, future greenhouse gas emissions will depend on political and economical decisions that are extremely difficult to predict decades in advance.

## References

Arrhenius, S. (1896). On the influence of carbonic acid in the air upon the temperature of the ground. Philosophical Magazine and Journal of Science, 41(5), 237–276.Find this resource:

BACC Author Team. (2008). Assessment of climate change for the Baltic Sea basin. Berlin: Springer Verlag.Find this resource:

BACC II Author Team. (2015). Second assessment of climate change for the Baltic Sea basin. London: Springer.Find this resource:

Bamber, J., & Riva, R. (2010). The sea level fingerprint of recent ice mass fluxes. The Cryosphere, 4, 621–627.Find this resource:

Benestad, R. (2016). Downscaling climate information. Oxford research encyclopedia on climate science.Find this resource:

Bindoff, N. L., Stott, P. A., AchutaRao, K. M., Allen, M. R., Gillett, N., Gutzler, D., . . . Zhang, X. (2013). Detection and attribution of climate change: From global to regional. In IPCC, Climate change 2013: The physical science basis (pp. 867–942). Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press.Find this resource:

Bjørge, D., & Haugen, J. E. (1998). Simulation of present-day climate in HIRHAM using perfect boundaries. RegClim General Technical Report, 1, 9–16.Find this resource:

Carter, T. R., Posch, M., & Tuomenvirta, H. (1996). The SILMU scenarios: Specifying Finland’s future climate for use in impact assessment. Geophysica, 32(1–2), 235–260.Find this resource:

Christensen, J. H., Carter, T. R., & Rummukainen, M. (2007). Evaluating the performance and utility of regional climate models: The PRUDENCE project. Climatic Change, 81, 1–6.Find this resource:

Christensen, O. B., Christensen, J. H., Machenhauer, B., & Botzet, M. (1998). Very high-resolution regional climate simulations over Scandinavia. Present climate. Journal of Climate, 11, 3204–3229.Find this resource:

Christensen, O. B., Kjellström, E., & Zorita, E. (2015). Projected change—Atmosphere. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 217–233). London: Springer.Find this resource:

Christensen, J. H., Räisänen, J., Iversen, T., Bjørge, D., Christensen, O. B., & Rummukainen, M. (2001). A synthesis of regional climate change simulations—A Scandinavian perspective. Geophysical Research Letters, 28(6), 1003–1006.Find this resource:

Church, J. A., Clark, P. U., Cazenave, A., Gregory, J. M., Jevrejeva, S., Levermann, A., . . . Unnikrishnan, A. S. (2013). Sea level change. In IPCC, Climate change 2013: The physical science basis (pp. 1137–1216). Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press.Find this resource:

Churkina, G., Zaehle, S., Hughes, J., Viovy, N., Chen, Y., Jung, M., . . . Jones, C. (2010). Interactions between nitrogen deposition, land cover conversion, and climate change determine the contemporary carbon balance of Europe. Biogeosciences, 7, 2749–2764.Find this resource:

Ciais, P., Sabine, C., Bala, G., Bopp, L., Brovkin, V., Canadell, J., . . . Thornton, P. (2013). Carbon and other biogeochemical cycles. In IPCC, Climate change 2013: The physical science basis (pp. 465–570). Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press.Find this resource:

Collins, M., Knutti, R., Arblaster, J., Dufresne, J.-L., Fichefet, T., Friedlingstein, P., . . . Wehner, M. (2013). Long-term climate change: Projections, commitments and irreversibility. In IPCC, Climate change 2013: The physical science basis (pp. 1029–1136). Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press.Find this resource:

DeConto, R. M., & Pollard, D. (2016). Contribution of Antarctica to past and future sea-level rise. Nature, 531(7596), 591–597.Find this resource:

Déqué, M., Somot, S., Sanchez-Gomez, E., Goodess, C. M., Jacob, D., Lenderink, G., & Christensen, O. B. (2012). The spread amongst ENSEMBLES regional scenarios: Regional climate models, driving general circulation models and interannual variability. Climate Dynamics, 38(5), 951–964.Find this resource:

Deser, C., Phillips, A., Bourdette, V., & Teng, H. (2012). Uncertainty in climate change projections: The role of internal variability. Climate Dynamics, 38(3), 527–546.Find this resource:

Dippner, J. W., Vuorinen, I., Daunys, D., Flinkman, J., Halkka, A., Köster, F. W., . . . Wasmund, N. (2008). Climate-related marine ecosystem change. In BACC Author Team, Assessment of climate change for the Baltic Sea basin (pp. 221–238). Berlin: Springer Verlag.Find this resource:

de Vries, W., & Posch, M. (2011). Modelling the impact of nitrogen deposition, climate change and nutrient limitations on tree carbon sequestration in Europe for the period 1900–2050. Environmental Pollution, 159(10), 2289–2299.Find this resource:

Eriksson, C., Omstedt, A., Overland, J. E., Percival, D. B., & Mofjeld, H. O. (2007). Characterizing the European sub-Arctic winter climate since 1500 using ice, temperature, and atmospheric circulation time series. Journal of Climate, 20(21), 5316–5334.Find this resource:

Graham, L. P., Andréasson, J., & Carlsson, B. (2007). Assessing climate change impacts on hydrology from an ensemble of regional climate models, model scales and linking methods—A case study on the Lule River basin. Climatic Change, 81(Suppl. 1), 293–307.Find this resource:

Graham, L. P., Chen, D., Christensen, O. B., Kjellström, E., Krysanova, V., Meier, H. E. M., . . . Ruosteenoja, K. (2008). Projections of future anthropogenic climate change. In BACC Author Team, Assessment of climate change for the Baltic Sea basin (pp. 133–219). Berlin: Springer Verlag.Find this resource:

Gräwe, U., & Burchard, H. (2012). Storm surges in the western Baltic Sea: The present and a possible future. Climate Dynamics, 39(1), 165–183.Find this resource:

Grinsted, A. (2015). Projected change—Sea level. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 253–263). London: Springer.Find this resource:

Grotch, S. L., & MacCracken, M. C. (1991). The use of general circulation models to predict regional climatic change. Journal of Climate, 4(3), 286–303.Find this resource:

Hägg, H. E., Humborg, C., Morth, C. M., Medina, M. R., & Wulff, F. (2010). Scenario analysis on protein consumption and climate change effects on riverine N export to the Baltic Sea. Environmental Science and Technology, 44(7), 2379–2385.Find this resource:

Hansen, J., Ruedy, R., Sato, M., & Lo, K. (2010). Global surface temperature change. Reviews of Geophysics, 48(4), RG4004.Find this resource:

Havenhand, J. N. (2012). How will acidification affect Baltic Sea ecosystems? An assessment of plausible impacts on key functional groups. Ambio, 41(6), 637–644.Find this resource:

Hawkins, E., & Sutton, R. (2009). The potential to narrow uncertainty in regional climate predictions. Bulletin of the American Meteorological Society, 90(8), 1095–1106.Find this resource:

Hawkins, E., & Sutton, R. (2011). The potential to narrow uncertainty in projections of regional precipitation change. Climate Dynamics, 37(1), 407–418.Find this resource:

HELCOM. (2009). Eutrophication in the Baltic Sea—An integrated thematic assessment of the effects of nutrient enrichment in the Baltic Sea region. Baltic Sea Environment Proceedings, 115B. Helsinki: Helsinki Commission.Find this resource:

HELCOM. (2011). The fifth Baltic Sea pollution load compilation (PLC-5). Baltic Sea Environment Proceedings, 128. Helsinki: Helsinki Commission.Find this resource:

HELCOM. (2015). Updated fifth Baltic Sea pollution load compilation (PLC-5.5). Baltic Sea Environment Proceedings, 145. Helsinki: Helsinki Commission.Find this resource:

Held, I. M., & Soden, B. J. (2006). Robust responses of the hydrological cycle to global warming. Journal of Climate, 19(21), 5686–5699.Find this resource:

Hill, E. M., Davis, J. L., Tamisiea, M. E., & Lidberg, M. (2010). Combination of geodetic observations and models for glacial isostatic adjustment fields in Fennoscandia. Journal of Geophysical Research, 115(B7), B07403.Find this resource:

Holmes, C. R., Woollings, T., Hawkins, E., & de Vries, H. (2016). Robust future changes in temperature variability under greenhouse gas forcing and the relationship with thermal advection. Journal of Climate, 29(6), 2221–2236.Find this resource:

Hong, B., Swaney, D. P., Morth, C. M., Smedberg, E., Hägg, H. E., Humborg, C., . . . Bouraoui, F. (2012). Evaluating regional variation of net anthropogenic nitrogen and phosphorus inputs (NANI/NAPI), major drivers, nutrient retention pattern and management implications in the multinational areas of Baltic Sea basin. Ecological Modelling, 227, 117–135.Find this resource:

Hordoir, R., & Meier, H. E. M. (2011). Effect of climate change on the thermal stratification of the Baltic Sea: A sensitivity experiment. Climate Dynamics, 38(9), 1703–1713.Find this resource:

Humborg, C., Andersen, H. E., Blenckner, T., Gadegast, M., Giesler, R., Hartmann, J., . . . Weyhenmeyer, G. (2015). Environmental impacts—Freshwater biogeochemistry. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 307–336). London: Springer.Find this resource:

IPCC (Intergovernmental Panel on Climate Change). (1990). Climate change. The IPCC scientific assessment. Cambridge, U.K.: Cambrige University Press.Find this resource:

IPCC (Intergovernmental Panel on Climate Change). (1996). Climate change 1995: The science of climate change. Contribution of Working Group I to the Second Assessment Report of the Intergovernmental Panel on Climate Change: Summary for Policymakers. Cambridge, U.K.: Cambrige University Press.Find this resource:

IPCC (Intergovernmental Panel on Climate Change). (2001). Climate change 2001: The scientific basis. Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change: Summary for Policymakers. Cambridge, U.K.: Cambridge University Press.Find this resource:

IPCC (Intergovernmental Panel on Climate Change). (2013). Climate change 2013: The physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press.Find this resource:

Jacob, D., Petersen, J., Eggert, N., Alias, A., Christensen, O. B., Bouwer, L. M., . . . Yiou, P. (2014). EURO-CORDEX: New high-resolution climate change projections for European impact research. Regional Environmental Change, 14(2), 563–578.Find this resource:

Jacob, D., & Podzun, R. (1997). Sensitivity studies with the regional climate model REMO. Meteorology and Atmospheric Physics, 63(1), 119–129.Find this resource:

Jensen, J., & Müller-Navarra, S. H. (2008). Storm surges on the German coast. Die Küste, 74 ICCE, 92–124.Find this resource:

Johansson, M. M., Pellikka, H., Kahma, K. K., & Ruosteenoja, K. (2014). Global sea level rise scenarios adapted to the Finnish coast. Journal of Marine Systems, 129(1), 35–46.Find this resource:

Jóhannesson, T., Jónsson T., Källén, E., & Kaas, E. (1995). Climate change scenarios for the Nordic countries. Climate Research, 5(3), 181–195.Find this resource:

Jönsson, A. M., Linderson, M.-L., Stjernquist, I., Schlyter P., & Bärring, L. (2004). Climate change and the effect of temperature backlashes causing frost damage in Picea abies. Global and Planetary Change, 44(1–4), 195–207.Find this resource:

Jylhä, K., Fronzek, S., Tuomenvirta, H., Carter, T. R., & Ruosteenoja, K. (2008). Changes in frost, snow and Baltic Sea ice by the end of the twenty-first century based on climate model projections for Europe. Climatic Change, 86(3), 441–462.Find this resource:

Kämäri, J. (1997). Preface. Boreal Environment Research, 2(1), 1.Find this resource:

Kjellström, E., Bärring, L., Jacob, D., Jones, R., Lenderink, G., & Schär, C. (2007). Modelling daily temperature extremes: Recent climate and future changes over Europe. Climatic Change, 81(S1), 249–265.Find this resource:

Knutti, R., Furrer, R., Tebaldi, C., Cermak, J., & Meehl, G. A. (2010). Challenges in combining projections from multiple climate models. Journal of Climate, 23(10), 2739–2758.Find this resource:

Koenigk, T., König Beatty, C., Caian, M., Döscher, R., & Wyser, K. (2012). Potential decadal predictability and its sensitivity to sea ice albedo parameterization in a global coupled model. Climate Dynamics, 38(11), 2389–2408.Find this resource:

Kullman, L. (2004). Long-term geobotanical observations of climate change impacts in the Scandes of West-Central Sweden. Nordic Journal of Botany, 24(4), 445–467.Find this resource:

Kullman, L. (2010). Alpine flora dynamics—A critical review of responses to climate change in the Swedish Scandes since the early 1950s. Nordic Journal of Botany, 28(4), 398–408.Find this resource:

Łabuz, T. A. (2015). Environmental impacts—Coastal erosion and coastline changes. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 381–396). London: Springer.Find this resource:

Langner, J., Andersson, C., & Engardt, M. (2009). Atmospheric input of nitrogen to the Baltic Sea basin: Present situation, variability due to meteorology and impact of climate change. Boreal Environment Research, 14(1), 226–237.Find this resource:

Lehtonen, I., Ruosteenoja, K., & Jylhä, K. (2014). Projected changes in European extreme precipitation indices on the basis of global and regional climate model ensembles. International Journal of Climatology, 34(4), 1208–1222.Find this resource:

Lindner, M., Maroschek, M., Netherer, S., Kremer, A., Barbati, A., Garcia-Gonzalo, J., . . . Marchetti, M. (2010). Climate change impacts, adaptive capacity, and vulnerability of European forest ecosystems. Forest Ecology and Management, 259(4), 698–709.Find this resource:

Luomaranta, A., Ruosteenoja, K., Jylhä, K., Gregow, H., Haapala, J., & Laaksonen, A. (2014). Multimodel estimates of the changes in the Baltic Sea ice cover during the present century. Tellus, 66A, 22617.Find this resource:

Maraun, D. (2013). When will trends in European mean and heavy daily precipitation emerge? Environmental Research Letters, 8(1), 014004.Find this resource:

Meier, H. E. M. (2006). Baltic Sea climate in the late twenty-first century: A dynamical downscaling approach using two global models and two emission scenarios. Climate Dynamics, 27(1), 39–68.Find this resource:

Meier, H. E. M. (2015). Projected change—Marine physics. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 243–251). London: Springer.Find this resource:

Meier, H. E. M., Andersson, H. C., Eilola, K., Gustafsson, B., Kuznetsov I., Müller-Karulis, B., . . . Savchuk, O. P. (2011). Hypoxia in future climates—A model ensemble study for the Baltic Sea. Geophysical Research Letters, 38(24), L24608.Find this resource:

Meier, H. E. M., Döscher, R., & Halkka, A. (2004). Simulated distributions of Baltic sea-ice in warming climate and consequences for the winter habitat of the Baltic ringed seal. Ambio, 33(4), 249–256.Find this resource:

Meier, H. E. M., Hordoir, R., Andersson, H. C., Dieterich, C., Eilola, K., Gustafsson, B. G., . . . Schimanke, S. (2012). Modeling the combined impact of changing climate and changing nutrient loads on the Baltic Sea environment in an ensemble of transient simulations for 1961–2099. Climate Dynamics, 39(9), 2421–2441.Find this resource:

Meier, H. E. M., Höglund, A., Eilola, K., & Almroth-Rosell, E. (2017). Impact of accelerated future global mean sea level rise on hypoxia in the Baltic Sea. Climate Dynamics, 49(1–2), 163–172.Find this resource:

Mitchell, J. F. B., Manabe, S., Meleshko, V., & Tokioka, T. (1990). Equilibrium climate change—And its implications for the future. In IPCC, Climate change. The IPCC scientific assessment (pp. 131–172). Cambridge, U.K.: Cambridge University Press.Find this resource:

Morice, C. P., Kennedy, J. J., Rayner, N. A., & Jones, P. D. (2012). Quantifying uncertainties in global and regional temperature change using an ensemble of observational estimates: The HadCRUT4 dataset. Journal of Geophysical Research, 117(D8), D08101.Find this resource:

Nakićenović, N., & Swart, R. (Eds.). (2000). Emissions scenarios. A special report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge, U.K.: Cambridge University Press.Find this resource:

Niemelä, P., Tolvanen, H., Rönkä, M., Kellomäki, S., Krug, J., Schurgers, G., . . . Kalliola, R. (2015). Environmental impacts—Coastal ecosystems, birds and forests. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 291–306). London: Springer.Find this resource:

Nikulin, G., Kjellström, E., Hansson, U., Jones, C., Strandberg, G., & Ullerstig, A. (2011). Evaluation and future projections of temperature, precipitation and wind extremes over Europe in an ensemble of regional climate simulations. Tellus, 63A(1), 41–55.Find this resource:

O’ Gorman, P. (2014). Contrasting responses of mean and extreme snowfall to climate change. Nature, 512(7515), 416–418.Find this resource:

Omstedt, A., Edman, M., Claremar, B., Frodin, P., Gustafsson, E., Humborg, C., . . . Yurova, A. (2012). Future changes of the Baltic Sea acid-base (pH) and oxygen balances. Tellus, 64B, 19586.Find this resource:

Omstedt, A., & Hansson, D. (2006). The Baltic Sea ocean climate system memory and response to changes in the water and heat balance components. Continental Shelf Research, 26(2), 236–251.Find this resource:

Orlowsky, B., & Seneviratne, S. I. (2012). Global changes in extreme events: Regional and seasonal dimension. Climatic Change, 116(3), 669–696.Find this resource:

Peltola, H., Kellomäki, S., & Väisänen, H. (1999). Model computations of the impact of climatic change on the windthrow risk of trees. Climatic Change, 41(1), 17–36.Find this resource:

Pryor, S. C., & Barthelmie, R. J. (2010). Climate change impacts on wind energy: A review. Renewable and Sustainable Energy Reviews, 14(1), 430–437.Find this resource:

Räisänen, J. (1994). A comparison of the results of seven GCM experiments in northern Europe. Geophysica, 30(1–2), 3–30.Find this resource:

Räisänen, J. (2002). CO2-induced changes in interannual temperature and precipitation variability in 19 CMIP2 experiments. Journal of Climate, 15(9), 2395–2411.Find this resource:

Räisänen, J. (2015). Twenty-first century changes in snowfall climate in Northern Europe in ENSEMBLES regional climate models. Climate Dynamics, 46(1), 339–353.Find this resource:

Räisänen, J., & Eklund, J. (2012). 21st century changes in snow climate in Northern Europe: A high-resolution view from ENSEMBLES regional climate models. Climate Dynamics, 38, 2575–2591.Find this resource:

Räisänen, J., Hansson, U., Ullerstig, A., Döscher, R., Graham, L. P., Jones, C., . . . Willén, U. (2004). European climate in the late 21st century: Regional simulations with two driving global models and two forcing scenarios. Climate Dynamics, 22(1), 13–31.Find this resource:

Räisänen, J., & Ruokolainen, L. (2006). Probabilistic forecasts of near-term climate change based on a resampling ensemble technique. Tellus, 58A(4), 461–472.Find this resource:

Räisänen, J., & Ylhäisi, J. S. (2015). CO2-induced climate change in northern Europe: CMIP2 versus CMIP3 versus CMIP5. Climate Dynamics, 45(7), 1877–1897.Find this resource:

Räty, O., Virta, H., Bosshard, T., & Connelly, C. (2017). Regional climate model and model output statistics method uncertainties and the effect of temperature and precipitation on future river discharges in Scandinavia. Hydrology Research,Find this resource:

Rummukainen, M., Räisänen, J., Bringfelt, B., Ullerstig, A., Omstedt, A., Willén, U., . . . Jones, C. (2000). A regional climate model for northern Europe—Model description and results from the downscaling of two GCM control simulations. Climate Dynamics, 17(5), 339–359.Find this resource:

Ruosteenoja, K., Jylhä, K., & Kämäräinen, M. (2016). Climate projections for Finland under the RCP forcing scenarios. Geophysica, 51(1), 17–50.Find this resource:

Rutgersson, A., Omstedt, A., & Räisänen, J. (2002). Net precipitation over the Baltic Sea during present and future climate conditions. Climate Research, 22(1), 27–39.Find this resource:

Sælthun, N. R. (1992). Klimaendringer og energiproduksjon: En orientering. Vannet i Norden [Newsletter of the Nordic Association for Hydrology], 25(2), 8–12.Find this resource:

Sælthun, N. R., Aittoniemi, P., Bergström, S., Einarsson, K., Jóhannesson, T., Lindström, G., . . . Aamodt, K. O. (1998). Climate change impacts on runoff and hydropower in the Nordic countries. TemaNord 1998: 522. Copenhagen: Nordic Council of Ministers.Find this resource:

Schimanke, S., Dieterich, C., & Meier, H. E. M. (2014). An algorithm based on sea-level pressure fluctuations to identify major Baltic inflow events. Tellus, 66A, 23452.Find this resource:

Schneider, B., Eilola, K., Lukkari, K., Muller-Karulis, B., & Neumann, T. (2015). Environmental impacts—Marine biogeochemistry. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 337–361). London: Springer.Find this resource:

Shepherd, T. G. (2014). Atmospheric circulation as a source of uncertainty in climate change projections. Nature Geoscience, 7(10), 703–708.Find this resource:

Simpson, D., Bartnicki, J., Jalkanen, J.-P., Hansson, H.-C., Hertel, O., Langner, J., & Pryor, S. C. (2015). Environmental impacts—Atmospheric chemistry. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 267–289). London: Springer.Find this resource:

Skjøth, C. A., & Geels, C. (2013). The effect of climate and climate change on ammonia emissions in Europe. Atmospheric Chemistry and Physics, 13(1), 117–128.Find this resource:

Slangen, A. B. A., Katsman, C., van de Wal, R. S. W., Vermeersen, L. L. A., & Riva R. E. M. (2012). Towards regional projections of twenty-first century sea-level change based on IPCC SRES scenarios. Climate Dynamics, 38(5), 1191–1209.Find this resource:

Smith, B., Aasa, A., Ahas, R., Blenckner, T., Callaghan, T. V., de Chazal, J., . . . Wolf, A. (2008). Climate-related change in terrestrial and freshwater ecosystems. In: BACC Author Team, Assessment of climate change for the Baltic Sea basin (pp. 221–238). Berlin: Springer Verlag.Find this resource:

Smith, T. M., Reynolds, R. W., Peterson, T. C., & Lawrimore, J. (2008). Improvements to NOAA’s historical merged land-ocean surface temperature analysis (1880–2006). Journal of Climate, 21(10), 2283–2296.Find this resource:

Sommer, U., & Lengfellner, K. (2008). Climate change and the timing, magnitude, and composition of the phytoplankton spring bloom. Global Change Biology, 14(6), 1199–1208.Find this resource:

Sonnenborg, T. O. (2015). Projected change—Hydrology. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 235–242). London: Springer.Find this resource:

Stouffer, R. J., Manabe, S., & Bryan, K. (1989). Interhemispheric asymmetry in climate response to a gradual increase of atmospheric CO2. Nature, 342, 660–662.Find this resource:

Sutton, M. A., Reis, S., Riddick, S. N., Dragosits, U., Nemitz, E., Theobald, M. R., . . . Vries, W. D. (2013). Towards a climate-dependent paradigm of ammonia emission and deposition. Philosophical Transactions of the Royal Society B: Biological Sciences, 368(1621), 20130166.Find this resource:

Taylor, K. E., Stouffer R. J., & Meehl, G. A. (2012). An overview of CMIP5 and the experiment design. Bulletin of the American Meteorological Society, 93(4), 485–498.Find this resource:

Tinz, B. (1996). On the relation between annual maximum extent of ice cover in the Baltic Sea level pressure as well as air temperature field. Geophysica, 32(3), 319–341.Find this resource:

Tuovinen, J.-P., Hakola, H., Karlsson, P. E., & Simpson, D. (2013). Air pollution risks to Northern European forests in a changing climate. In R. Matyssek, N. Clarke, P. Cudlin, T. N. Mikkelsen, J. P. Tuovinen, G. Wieser, & E. Paoletti (Eds.), Climate change, air pollution and global challenges: Understanding and perspectives from forest research (pp. 77–99). Amsterdam: Elsevier.Find this resource:

van der Linden, P., & Mitchell, J. F. B. (2009). ENSEMBLES: Climate change and its impacts: Summary of research and results from the ENSEMBLES project. Retrieved from http://ensembles-eu.metoffice.com/docs/Ensembles_final_report_Nov09.pdf.

van Vuuren, D. P., Edmons, J., Kainuma, M., Riahi, K., Thomson, A., Hibbard, K., . . . Rose, S. K. (2011). The representative concentration pathways: An overview. Climatic Change, 109(1–2), 5–31.Find this resource:

Vehviläinen, B., & Huttunen, M. (1997). Climate change and water resources in Finland. Boreal Environment Research, 2(1), 3–18.Find this resource:

Veijalainen, N., Dubrovin, T., Marttunen, M., & Vehviläinen, B. (2010). Climate change impacts on water resources and lake regulation in the Vuoksi watershed in Finland. Water Resources Management, 24(13), 3437–3459.Find this resource:

Veraart, A. J., de Klein, J. J. M., & Scheffer, M. (2011). Warming can boost denitrification disproportionately due to altered oxygen dynamics. PLoS ONE, 6(3), e18508.Find this resource:

Viitasalo, M., Blenckner, T., Gårdmark, A., Kaartokallio, H., Kautsky, L., Kuosa, H., . . . Wikner, J. (2015). Environmental impacts—Marine ecosystems. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 363–380). London: Springer.Find this resource:

Vousdoukas, M. I., Voukouvalas, E., Annunziato, A., Giardino, A., & Feyen, L. (2016). Projections of extreme storm surge levels along Europe. Climate Dynamics, 47(9–10), 3171–3190.Find this resource:

Watson, R. T., Rodhe, H., Oeschger, H., & Siegenthaler, U. (1990). Greenhouse gases and aerosols. In IPCC, Climate change. The IPCC scientific assessment (pp. 1–40). Cambridge, U.K.: Cambridge University Press.Find this resource:

Wibig, J., Maraun, D., Benestad, R., Kjellström, E., Lorenz, P., & Christensen, O. B. (2015). Projected change—Models and methodology. In BACC II Author Team, Second assessment of climate change for the Baltic Sea basin (pp. 189–215). London: Springer.Find this resource:

Ylhäisi, J. S., & Räisänen, J. (2014). Twenty-first century changes in daily temperature variability in CMIP3 climate models. International Journal of Climatology, 34(5), 1414–1428.Find this resource:

Zaehle, S., Bondeau, A., Carter, T. R., Cramer, W., Erhard, M., Prentice, I. C., . . . Sykes, M. (2007). Projected changes in terrestrial carbon storage in Europe under climate and land-use change, 1990–2100. Ecosystems, 10(3), 380–401.Find this resource: