Transcript Document

Comparative analysis of climatic variability
characteristics of the Svalbard archipelago and the
North European region based on meteorological
stations network data
Daria Vasilyeva (St-Petersburg State University)
Project : “Meteo-glaciological monitoring of massheat exchange of glaciers”
Supervisors: Pavel Svyashchennikov (AARI),
Jack Kohler (NPI)
THE GOAL:
Studying of the climate change in the Svalbard archipelago
and the North European region
Focal points:
 To reveal climate change tendencies in 1930 – 2003
and in 1993 – 2003 in different seasons
 To analyze spatial climatic variability structure
 To estimate long term oscillations contribution in climatic
variability
 To describe climate regimes in the Atlantic sector of the Arctic
Data
Location of meteorological stations in the study area:
•, • - analyzable stations over the period 1930 – 2003
• - additional analyzable stations over the period 1993 – 2003
Methods
1.
 positive monthly average surface air temperature sums (PMASATS)
 negative monthly average surface air temperature sums (NMASATS)
sum of positive monthly average surface air temperatures
of one year is sum of all positive monthly average values of year:
(t>0)= ti(>0), where
(t>0) – PMASATS,
ti(>0) – monthly average positive temperature
sum of negative monthly average surface air temperatures
of one year is sum of all negative monthly average values,
beginning from autumn of previous year, i. e. for example the sum of
negative temperatures over 1931 is accumulated from negative values of
monthly temperature from November till December 1930 and from
January till May 1931:
(t≤0)= ti(≤0), where
(t≤0) – NMASATS,
ti(≤0) – monthly average negative temperature
Methods
Reasons:
Positive air temperature sum can be considered as a value, proportional heat
of ice and snow fusion.
Negative air temperature sum can be considered as a value, which
determines cold content.
Model results had shown (Makshtas et al, 2003), that ice cover is extremely
sensitive to positive temperature changing.
From the point of statistical analysis using such value as temperature sum
allows to weaken weather, in this case noise, component
(Alekseev, Svyaschennikov, 1991).
Usage of such characteristic is convenient also as variance of values sum
equal variances sum of these values, thus if we use temperature sum, then
we receive value, having larger variability in comparison with monthly
temperature. It is convenient to use more variable characteristic to reveal
climate changes.
Methods
2.
Method of cores (or delta-like functions) were used for calculation of
probability density functions of PMASATS and NMASATS.
Reason:
This method allows to find trusty assessments of probability density in
sufficiently short set of observations. Dispersion of assessment of probability
density function is several times shorter than variance of assessment were
found with more prevalent histograms method. The histograms method is
not correct for short time series (Alekseev, Svyaschennikov, 1991).
Trends of PMASATS during the period 1930 – 2003 (black marked numbers are
values (0C) of not statistically significant trends, red marked numbers are values
(0C) of significant trends (significance level less 0.05):
• - positive trends
• - negative trends
• - no trends
Trends of NMASATS during the period 1930 – 2003 (black marked numbers are values
(0C) of not statistically significant trends, red marked numbers are values (0C) of
significant trends (significance level less 0.05):
• - positive trends
• - negative trends
• - no trends
Trends of PMASATS during the period 1993 – 2003 (black marked numbers are values
(0C) of not statistically significant trends, red marked numbers are values (0C) of
significant trends (significance level less 0.05):
• - positive trends, basic stations
• - positive trends, additional stations
• - negative trends, basic stations,
• - negative trends, additional stations
Trends of NMASATS during the period 1993 – 2003 (black numbers are values
(0C) of not statistically significant trends, red numbers are values (0C) of
significant trends (significance level less 0.05):
• - positive trends, basic stations
• - positive trends, additional stations
• - negative trends, basic stations
• - negative trends, additional stations
Middle of warmest periods (maxima of PMASATS):
• - 1935 – beginning of 1940s
• - end of 1950s
• - end of 1980s – end1990s
Middle of the warmest periods (Maxima of NMASATS):
• - beginning of 1940s
• - 1949, 1955
• - 1990s
Contribution of long term (fraction unit) oscillations into variance of
PMASATS (oscillation period is more than 12 years) during the
period 1930 – 2003
Contribution of long term (fraction unit) oscillations into variance of
NMASATS (oscillation period is more than 12 years) during the period
1930 – 2003
P (
0.1
b
P (
a
0.05
0.08
0.04
0.06
0.03
0.04
0.02
0.02
0.01
0
0
-15
-10
-5
0
5
10
15
-30
-20
-10
0
10
20
'
Probability density function of: a) PMASATS (Murmansk), b) NMASATS
(Bjornoya island)
30
'
Spatial distribution of PMASATS probability density functions types:
- single-modal distribution
- bimodal distribution
Spatial distribution of NMASATS probability density functions types:
- single-modal distribution
- bimodal distribution
Pressure field of July – August in 1953 (warm year)
Pressure field of July – August in 1966 (cold year)
Pressure field of January –February in 1954 (warm year)
Pressure field of January –February in 1966 (cold year)
Conclusions:
As a whole our investigations had shown, that usage of PMASATS and NMASATS
was convenient to determine the two seasons in the Arctic, justified and
reasonable.
Our results as well as results of other researchers evidence the complex nature
of climate change during measurements period in the Atlantic sector of the Arctic
and it can not be brought to anthropogenic impact only.
Climatic variability study in measurements period from 1930 to 2003 had
shown, that positive tendency of PMASATS predominates in the region in whole.
Overall cooling is observed in cold season. However trends are statistically
significant far from all. But question of unidirectional tendencies chance for the
most station requires more investigation in detail.
Modern time (1993 – 2003) is characterized with warming in general.
Three warming periods over 1930 – 2003 were distinguished, especially in
warm period. The means of sums temperatures maxima of the periods turned out
very close. These periods are 1935 – beginning of 1940s, the end of 1950s and
the end of 1980s – end of 1990s for PMASATS and the beginning of 1940s;
1949,1955; 1990s for NMASATS, that reveal warming beginning in warm season
before cold season.
Conclusions:
The contribution of long-term oscillations into dispersion of PMASATS
decreases in general eastward in the study region. The contribution of
long-term oscillations into the dispersion of NMASATS is characterized with
values decreasing southward and eastward in the study region.
Nonuniqueness of climate regime of study area investigation had shown,
that some stations had single-modal distribution, others had bimodal one.
We can interpret such bimodal distribution as two climate regimes
presence. Our results of bimodal distribution of probability density function
of temperatures sums presence evidences that mean and variance are not
sufficient to climate regime description of study area. Information of
probability density function requires for the complete climate distribution.
As a whole the obtained results evidence that in spite of global warming
Arctic regional climate changes are complex. There are short-term
oscillations and internal dynamical factor can be cause of climate change
without external factors.