Zgryźliwość kojarzy mi się z radością, która źle skończyła.
Chapter 16
Changes in Sea Surface Temperature
of the South Baltic Sea (1854–2005)
Andrzej A. Marsz and Anna Styszyńska
16.1 Stating the Problem
There have been hundreds of works written on the changes in the air temperature
over Poland and neighboring countries. These changes, at least, during instrumental
observations are well known. The literature dealing with the changes in sea surface
temperature of the Baltic Sea is relatively poor. Soskin
(1963)
analyses the changes
in sea surface temperature (SST) in the period 1900–1950 and notes in the 20s–30s
of the twentieth century the increase in temperature in relation to the preceding
period. Betin and Preobraženskij
(1962)
while dealing with the severe character of
winters in Europe refer to a series of information about the presence of ice cover in
the Baltic Sea and its duration (tenth–eighteenth centuries). This, in an indirect way,
gives some information regarding many centuries’ changes in winter water tem-
perature. These data however, are not continuous as they base on historic docu-
ments (chronicles, diaries, and harbour, merchant and customs documents) and
enable to derive only very general conclusions, regarding changes in the tempera-
ture of waters of the Baltic Sea, limited solely to winter periods.
Numerous remarks about changes in SST over short periods and in small sea
areas can be found in works dealing with biological oceanography and ecology of
the Baltic Sea; they concern different periods after the 1960s. Even, the impressive
in size, hydro meteorological monograph of the Baltic Sea (Terziev et al.
1992)
,
except for a map illustrating distribution of SST, does not deal with many-year
changes in SST. Some Polish monographs on coastal climate or on Polish coastal
zone mention changes in SST (e.g. Miętus et al.
2004)
in the off shore area.
Systematic measures of water temperature at measuring points of IMGW,
1
which in
A.A. Marsz and A. Styszyńska
Department of Meteorology and Nautical Oceanography, Gdynia Maritime University,
Sędzickiego 19, 81-374, Gdynia Nautical
e-mail: aamarsz@am.gdynia.pl
e-mail: stysa@am.gdynia.pl
1
Institute of Meteorology and Water Management in Poland.
R. Przybylak et al. (eds.),
The Polish Climate in the European Context:
An Historical Overview
, DOI 10.1007/978-90-481-3167-9_16,
© Springer Science + Business Media B.V. 2010
355
356
A.A. Marsz and A. Styszyńska
most cases are located in port waters and the very reading is done close to the shore
or in a distance of a few meters from the shore, provide the data.
In 2003 the authors (Marsz and Styszyńska
2003)
making use of the data from
COADS
2
presented changes in SST in the sea area covering the Gdańsk Bay and
the Gdańsk Deep in the years 1871–1992. They stated that there is statistically
significant positive trend in SST in this sea area (+0.009°C year
−1
, p < 0.005) and
strong correlation between changes in SST and the character of winter atmospheric
circulation observed in the examined period. Zblewski
(2006)
carried out a detailed
analysis of changes in SST in the whole Baltic Sea in the period 1982–2002, in
which very strong increase in air temperature was observed over the Baltic Sea and
in regions adjacent to the Baltic Sea. The aim of this work was to find out how the
strong warming of the atmosphere influences SST. The author noted that strong
positive trends, in most cases are statistically significant and what is more, indicate
clear seasonal variability in space almost in the entire surface of the Baltic Sea. The
annual trends in SST defined by Zblewski turned out to be much stronger than those
noted by the authors in the many- year period 1971–1992. Siegel et al.
(2006)
ana-
lyzed changes in SST of the Baltic Sea from Arkona Deep to the end of the Bothnia
Bay over the period 1990–2004. The conclusions they have arrived at, are, to a great
extent, similar to the results obtained by Zblewski
(2006)
.
The most recent works on changes in SST in the Baltic Sea were published in 2008.
Assessment of Climate Change for the Baltic Sea Basin
(2008
; later referred as
ACCBSB) presents the results of modeling of changes in heat amount in the Baltic Sea
and its regions which were observed in 1958–2005 and 1970–2005. As it can be seen
in the results presented by ACCBSB (2008; Fig. 2.49) a visible increase in the heat
amount in the Baltic took place in 1958–2005. Hansson and Omstedt
(2008)
basing on
the data from the twentieth century reconstructed the SST course and Maximum Ice
Extent (MIE) for the last 500 years. The above mentioned results indicate that in the
twentieth century SST was higher than in the last 500-year period and that the highest
decadal values of SST were observed in the 1930s and in the 1730s. The changes in
SST and MIE in the Baltic are within the limits of natural climate variability.
Changes in SST in open waters of the Baltic Sea,
3
because of the presence of a
specific for this sea density stratification, occur only under the influence of local ele-
ments responsible for climate formation. The heat resources transported into the
Baltic Sea with waters flowing from the North Sea have no contact with the sea sur-
face and that is why the processes of heat advection with the transported waters are
completely neglected for changes in SST. In the same way, changes in the sea surface
caused by human activity are neglected. Such activities performed on land by chang-
ing the way the land is used, changes in its moisture, forming city islands of warmth
may have influence on the temperature of ground and on the air temperature.
2
COADS – Comprehensive Ocean-Atmosphere Data Set.
3
Open, that is, situated in a certain distance from the shore, outside the area being under the influ-
ence of processes active in the coastal zone, where the local, especially in the sea areas close to
the port and in the regions in the vicinity of river estuaries anthropogenic and natural deformations
in the course of SST can be observed. This work completely neglects problems of changes in SST
in coastal and sheltered regions, dealing only with changes present in open waters.
16 Changes in Sea Surface Temperature of the South Baltic Sea (1854–2005)
357
Changes in SST are influenced by annual heat balance. On the side of heat gain
in the sea surface the only element that matters is the gain of solar radiation and
atmospheric re-radiation. On the side of loss there is radiation from the sea surface
and heat flux from the sea surface to the atmosphere. The latter is made up of sen-
sible heat flux (turbulent exchange) and of latent heat flux (latent heat of evapora-
tion). The values both of the streams of heat gain, as well as, heat losses are
influenced by changes in weather phenomena both periodically and aperiodically.
Because of great heat volume of water and large masses of water and at the same
time great thermal inertion of the layer of the Baltic waters above halocline, SST
‘records’ in its course rhythm of changes in weather conditions observed over lon-
ger periods and at the same time with different scale of delays, influences the
course of these conditions. Taking into consideration the above, it can be stated that
changes in SST of the Baltic represent resultant of the changes in regional climatic
conditions over the examined period and are free of anthropogenic influence.
4
The aim of this work is to present the course of changes in annual SST in the
southern part of the Baltic Sea, observed over the period of the past 152 years, that
is in the period from 1854 to 2005. The analysis of the course of changes in SST of
the Baltic Sea carried out for a longer period can solve a lot of problems and the
ones which seem to be most important, that is defining the scale of changes in SST,
defining the cooling and warming periods observed in the sea surface of the Baltic
Sea, defining the concordance of changes in the course of SST and the air tempera-
ture on land in the vicinity of the examined sea area and explaining what climatic
signal is indicated by changes in SST.
16.2 Data
The basic data were made up of chronological series of monthly values of SST from
the data set ER SST v.2.
5
This set contains global values of monthly SST which are
average values for areas 2°f × 2°l, with evenly nominated central points of these
areas (grid organization). The set ERSST v.2 for the period 1854–1992 is transformed
from COADS SST data, for the later period – high resolution satellite data, calibrated
by measurements in situ. How this set is constructed and what techniques are used to
get rid of interference, how the mean values and how the climatologic homogeneity
are obtained, can be found in works by Smith and Reynolds
(2004)
. The data from
this set are less accurate in the preliminary period and from both world wars because
of not equal number of data used for estimating mean values.
4
The only anthropogenic factor which has influence on changes in SST of the Baltic Sea is the change
in the concentration of CO
2
in the atmosphere. This results in changes in elements of the radiation
balance. The changes in CO
2
concentration are global so changes of the elements of the radiation bal-
ance over the Baltic should be the same as over the area adjacent to this sea.
5
The full name of the data set NOAA NCDC ERSST version2 is improved extended reconstructed
global sea surface temperature data based on COADS data.
358
A.A. Marsz and A. Styszyńska
The analysis of changes in SST in the Baltic Sea made use of a grid with coordinates
56°N, 18°E whose time series describes the mean SST defined within the limits
55–57°N, 17–19°E. The surface area of the sea area calculated as a flat area is
27,618 km
2
. Figure
16.1
presents the location of this surface. The described sea area
almost in 100% covers water surface and characterizes open waters of the southern
part of the Baltic Sea.
The standard estimation error for the mean monthly SST in the examined sea
area in most cases is within the range from ±0.01 to ±0.04°C, maximum errors
reach ±0.61°C (data set NOAA NCDC ERSST version2 err). Figure
16.2
presents
the distribution of estimation error for annual SST calculated as mean value of
monthly errors in a given year. The highest values of standard estimation errors for
monthly temperature, except for single cases, are noted in April.
The values of annual temperatures used for this analysis were calculated from
the values of mean monthly temperatures as mean arithmetic values. Changes in
annual SST in this grid point are very strongly correlated (r = 0.97–0.99)
6
with
Fig. 16.1
The location of areas whose mean annual temperatures were analysed in this work.
Grid 56°N, 18°E is marked with
black point
6
r – Pearson’s linear correlation coefficient.
16 Changes in Sea Surface Temperature of the South Baltic Sea (1854–2005)
359
Fig. 16.2
Distribution in time of standard errors of estimation of mean annual SST in grid
56°N, 18°E
the changes in SST in the adjacent to the examined grid points and this makes it
possible to state that they are representative for a far greater sea area than the
examined surface.
The data showing the air temperature from Stockholm station up to 1889 are
derived from the data set GHCN v.2 (Peterson and Vose
1997)
and for the year
1890 from the data set Nordklim (Tuomenvirta et al.
2001)
The data character-
izing the temperature at Hel till the year 1995 are taken from the work by
Miętus
(1998)
and in the following years they were supplemented with official
data from IMGW. The quality of these data has been checked by the authors of
these series and they are homogeneous. The values of NAO indexes used in this
work are taken from the data set accessible in official web sites WWW CRU
and J. Hurrell.
This work made use of standard methods in statistical analysis; when analyzing
signals a standard analytical methodology of electrical courses was employed
(Osiowski and Szabatin
1995)
. The principle of this method is that the following
elements are analyzed one by one, that is the course of deviation from the mean
value, low and up band signal envelopes whose aim is to define the components of
modulation, spectral analysis of a signal whose aim is to define spectrum of modu-
lating harmonic and harmonic being beating-up of modulating signals
7
and identi-
fication of impulse interference.
7
In case when two (or more) signals are received in the summing up system, processes of beat-
ing up (mixing) of signals forming new harmonics are observed. The basic harmonics of beating
up are the sum and difference between primary frequencies. In case when certain phase shifts
between primary signals are present, the amplitude of beating up harmonics can be higher than
the amplitude of modulating signals. The summing up system in this case is the surface layer of
the sea.
zanotowane.pl doc.pisz.pl pdf.pisz.pl hannaeva.xlx.pl
Changes in Sea Surface Temperature
of the South Baltic Sea (1854–2005)
Andrzej A. Marsz and Anna Styszyńska
16.1 Stating the Problem
There have been hundreds of works written on the changes in the air temperature
over Poland and neighboring countries. These changes, at least, during instrumental
observations are well known. The literature dealing with the changes in sea surface
temperature of the Baltic Sea is relatively poor. Soskin
(1963)
analyses the changes
in sea surface temperature (SST) in the period 1900–1950 and notes in the 20s–30s
of the twentieth century the increase in temperature in relation to the preceding
period. Betin and Preobraženskij
(1962)
while dealing with the severe character of
winters in Europe refer to a series of information about the presence of ice cover in
the Baltic Sea and its duration (tenth–eighteenth centuries). This, in an indirect way,
gives some information regarding many centuries’ changes in winter water tem-
perature. These data however, are not continuous as they base on historic docu-
ments (chronicles, diaries, and harbour, merchant and customs documents) and
enable to derive only very general conclusions, regarding changes in the tempera-
ture of waters of the Baltic Sea, limited solely to winter periods.
Numerous remarks about changes in SST over short periods and in small sea
areas can be found in works dealing with biological oceanography and ecology of
the Baltic Sea; they concern different periods after the 1960s. Even, the impressive
in size, hydro meteorological monograph of the Baltic Sea (Terziev et al.
1992)
,
except for a map illustrating distribution of SST, does not deal with many-year
changes in SST. Some Polish monographs on coastal climate or on Polish coastal
zone mention changes in SST (e.g. Miętus et al.
2004)
in the off shore area.
Systematic measures of water temperature at measuring points of IMGW,
1
which in
A.A. Marsz and A. Styszyńska
Department of Meteorology and Nautical Oceanography, Gdynia Maritime University,
Sędzickiego 19, 81-374, Gdynia Nautical
e-mail: aamarsz@am.gdynia.pl
e-mail: stysa@am.gdynia.pl
1
Institute of Meteorology and Water Management in Poland.
R. Przybylak et al. (eds.),
The Polish Climate in the European Context:
An Historical Overview
, DOI 10.1007/978-90-481-3167-9_16,
© Springer Science + Business Media B.V. 2010
355
356
A.A. Marsz and A. Styszyńska
most cases are located in port waters and the very reading is done close to the shore
or in a distance of a few meters from the shore, provide the data.
In 2003 the authors (Marsz and Styszyńska
2003)
making use of the data from
COADS
2
presented changes in SST in the sea area covering the Gdańsk Bay and
the Gdańsk Deep in the years 1871–1992. They stated that there is statistically
significant positive trend in SST in this sea area (+0.009°C year
−1
, p < 0.005) and
strong correlation between changes in SST and the character of winter atmospheric
circulation observed in the examined period. Zblewski
(2006)
carried out a detailed
analysis of changes in SST in the whole Baltic Sea in the period 1982–2002, in
which very strong increase in air temperature was observed over the Baltic Sea and
in regions adjacent to the Baltic Sea. The aim of this work was to find out how the
strong warming of the atmosphere influences SST. The author noted that strong
positive trends, in most cases are statistically significant and what is more, indicate
clear seasonal variability in space almost in the entire surface of the Baltic Sea. The
annual trends in SST defined by Zblewski turned out to be much stronger than those
noted by the authors in the many- year period 1971–1992. Siegel et al.
(2006)
ana-
lyzed changes in SST of the Baltic Sea from Arkona Deep to the end of the Bothnia
Bay over the period 1990–2004. The conclusions they have arrived at, are, to a great
extent, similar to the results obtained by Zblewski
(2006)
.
The most recent works on changes in SST in the Baltic Sea were published in 2008.
Assessment of Climate Change for the Baltic Sea Basin
(2008
; later referred as
ACCBSB) presents the results of modeling of changes in heat amount in the Baltic Sea
and its regions which were observed in 1958–2005 and 1970–2005. As it can be seen
in the results presented by ACCBSB (2008; Fig. 2.49) a visible increase in the heat
amount in the Baltic took place in 1958–2005. Hansson and Omstedt
(2008)
basing on
the data from the twentieth century reconstructed the SST course and Maximum Ice
Extent (MIE) for the last 500 years. The above mentioned results indicate that in the
twentieth century SST was higher than in the last 500-year period and that the highest
decadal values of SST were observed in the 1930s and in the 1730s. The changes in
SST and MIE in the Baltic are within the limits of natural climate variability.
Changes in SST in open waters of the Baltic Sea,
3
because of the presence of a
specific for this sea density stratification, occur only under the influence of local ele-
ments responsible for climate formation. The heat resources transported into the
Baltic Sea with waters flowing from the North Sea have no contact with the sea sur-
face and that is why the processes of heat advection with the transported waters are
completely neglected for changes in SST. In the same way, changes in the sea surface
caused by human activity are neglected. Such activities performed on land by chang-
ing the way the land is used, changes in its moisture, forming city islands of warmth
may have influence on the temperature of ground and on the air temperature.
2
COADS – Comprehensive Ocean-Atmosphere Data Set.
3
Open, that is, situated in a certain distance from the shore, outside the area being under the influ-
ence of processes active in the coastal zone, where the local, especially in the sea areas close to
the port and in the regions in the vicinity of river estuaries anthropogenic and natural deformations
in the course of SST can be observed. This work completely neglects problems of changes in SST
in coastal and sheltered regions, dealing only with changes present in open waters.
16 Changes in Sea Surface Temperature of the South Baltic Sea (1854–2005)
357
Changes in SST are influenced by annual heat balance. On the side of heat gain
in the sea surface the only element that matters is the gain of solar radiation and
atmospheric re-radiation. On the side of loss there is radiation from the sea surface
and heat flux from the sea surface to the atmosphere. The latter is made up of sen-
sible heat flux (turbulent exchange) and of latent heat flux (latent heat of evapora-
tion). The values both of the streams of heat gain, as well as, heat losses are
influenced by changes in weather phenomena both periodically and aperiodically.
Because of great heat volume of water and large masses of water and at the same
time great thermal inertion of the layer of the Baltic waters above halocline, SST
‘records’ in its course rhythm of changes in weather conditions observed over lon-
ger periods and at the same time with different scale of delays, influences the
course of these conditions. Taking into consideration the above, it can be stated that
changes in SST of the Baltic represent resultant of the changes in regional climatic
conditions over the examined period and are free of anthropogenic influence.
4
The aim of this work is to present the course of changes in annual SST in the
southern part of the Baltic Sea, observed over the period of the past 152 years, that
is in the period from 1854 to 2005. The analysis of the course of changes in SST of
the Baltic Sea carried out for a longer period can solve a lot of problems and the
ones which seem to be most important, that is defining the scale of changes in SST,
defining the cooling and warming periods observed in the sea surface of the Baltic
Sea, defining the concordance of changes in the course of SST and the air tempera-
ture on land in the vicinity of the examined sea area and explaining what climatic
signal is indicated by changes in SST.
16.2 Data
The basic data were made up of chronological series of monthly values of SST from
the data set ER SST v.2.
5
This set contains global values of monthly SST which are
average values for areas 2°f × 2°l, with evenly nominated central points of these
areas (grid organization). The set ERSST v.2 for the period 1854–1992 is transformed
from COADS SST data, for the later period – high resolution satellite data, calibrated
by measurements in situ. How this set is constructed and what techniques are used to
get rid of interference, how the mean values and how the climatologic homogeneity
are obtained, can be found in works by Smith and Reynolds
(2004)
. The data from
this set are less accurate in the preliminary period and from both world wars because
of not equal number of data used for estimating mean values.
4
The only anthropogenic factor which has influence on changes in SST of the Baltic Sea is the change
in the concentration of CO
2
in the atmosphere. This results in changes in elements of the radiation
balance. The changes in CO
2
concentration are global so changes of the elements of the radiation bal-
ance over the Baltic should be the same as over the area adjacent to this sea.
5
The full name of the data set NOAA NCDC ERSST version2 is improved extended reconstructed
global sea surface temperature data based on COADS data.
358
A.A. Marsz and A. Styszyńska
The analysis of changes in SST in the Baltic Sea made use of a grid with coordinates
56°N, 18°E whose time series describes the mean SST defined within the limits
55–57°N, 17–19°E. The surface area of the sea area calculated as a flat area is
27,618 km
2
. Figure
16.1
presents the location of this surface. The described sea area
almost in 100% covers water surface and characterizes open waters of the southern
part of the Baltic Sea.
The standard estimation error for the mean monthly SST in the examined sea
area in most cases is within the range from ±0.01 to ±0.04°C, maximum errors
reach ±0.61°C (data set NOAA NCDC ERSST version2 err). Figure
16.2
presents
the distribution of estimation error for annual SST calculated as mean value of
monthly errors in a given year. The highest values of standard estimation errors for
monthly temperature, except for single cases, are noted in April.
The values of annual temperatures used for this analysis were calculated from
the values of mean monthly temperatures as mean arithmetic values. Changes in
annual SST in this grid point are very strongly correlated (r = 0.97–0.99)
6
with
Fig. 16.1
The location of areas whose mean annual temperatures were analysed in this work.
Grid 56°N, 18°E is marked with
black point
6
r – Pearson’s linear correlation coefficient.
16 Changes in Sea Surface Temperature of the South Baltic Sea (1854–2005)
359
Fig. 16.2
Distribution in time of standard errors of estimation of mean annual SST in grid
56°N, 18°E
the changes in SST in the adjacent to the examined grid points and this makes it
possible to state that they are representative for a far greater sea area than the
examined surface.
The data showing the air temperature from Stockholm station up to 1889 are
derived from the data set GHCN v.2 (Peterson and Vose
1997)
and for the year
1890 from the data set Nordklim (Tuomenvirta et al.
2001)
The data character-
izing the temperature at Hel till the year 1995 are taken from the work by
Miętus
(1998)
and in the following years they were supplemented with official
data from IMGW. The quality of these data has been checked by the authors of
these series and they are homogeneous. The values of NAO indexes used in this
work are taken from the data set accessible in official web sites WWW CRU
and J. Hurrell.
This work made use of standard methods in statistical analysis; when analyzing
signals a standard analytical methodology of electrical courses was employed
(Osiowski and Szabatin
1995)
. The principle of this method is that the following
elements are analyzed one by one, that is the course of deviation from the mean
value, low and up band signal envelopes whose aim is to define the components of
modulation, spectral analysis of a signal whose aim is to define spectrum of modu-
lating harmonic and harmonic being beating-up of modulating signals
7
and identi-
fication of impulse interference.
7
In case when two (or more) signals are received in the summing up system, processes of beat-
ing up (mixing) of signals forming new harmonics are observed. The basic harmonics of beating
up are the sum and difference between primary frequencies. In case when certain phase shifts
between primary signals are present, the amplitude of beating up harmonics can be higher than
the amplitude of modulating signals. The summing up system in this case is the surface layer of
the sea.