Transcript Document

Congresso del Dipartimento di Fisica
Highlights in Physics 2005
11–14 October 2005, Dipartimento di Fisica, Università di Milano
CLIMATE CHANGE AND VARIABILITY IN ITALY IN THE LAST TWO CENTURIES
Brunetti M.*, Buffoni L.**, Lentini G.†, Mangianti F.††, Maugeri M.†, Monti F.†, Nanni T.*, Pastorelli R .†
*ISAC-CNR – Via Gobetti 101, I-40129 Bologna, Italy
**Osservatorio Astronomico di Brera – Via Brera 28, I-20121 Milan, Italy
†Istituto di Fisica Generale Applicata - Via Brera 28, I-20121 Milan, Italy
††Ufficio Centrale di Meteorologia Agraria – Via del Caravita 7A, I-00186 Rome, Italy
The awareness of the importance of data quality and homogeneity issues for a correct detection of climate change has increased rapidly in the last few years. Considering the results obtained within the EU-funded ALPCLIM project (see Böhm
et al., 2001), in the year 2000 the authors joined the CLIMAGRI research programme with the aim of better investigating the impact of data quality and homogeneity issues on the detection of Italian temperature and precipitation trends over the
last two centuries. The final goal was to extend, revise, improve and update the data-sets of Italian monthly secular temperature and precipitation series (Maugeri and Nanni, 1998; Buffoni et al., 1999; Brunetti et al., 2000) and to give more
reliable long-term trend estimates. The poster displays a synthesis of the principal results of the research program. Full details are discussed in Brunetti et al., 2005.
Italian Temperature Series
Italian Precipitation Series
INTRODUCTION
Around the mid 1990s the authors set up a wide research program with the aim of getting a better understanding of
the evolution of Italian climate in the last 150/200 years. This program allowed to obtain a first set of thermometric
and pluviometric records that was discussed in Maugeri and Nanni, 1998, in Buffoni et al., 1999 and in Brunetti et al.,
2000. Thanks to the CLIMAGRI project and to other international projects, in the year 2000 a new research program
was set up with the aim of constructing a new database with a great improvement of station density and metadata
availability.
Monthly precipitation and temperature series included in the new data-sets are shown and information on data
availability are also given.
Development of the final homogenised temperature and precipitation
networks.
TEMPERATURE PCA/CLUSTERING
METHODS
PRECIPITATION PCA/CLUSTERING
QUALITY CHECK AND MISSING DATA FILLING
All daily records were quality checked by means of crossed control and individual analysis
of all the values that markedly disagreed with the ones of surrounding stations. Then,
some minor gaps in precipitation records were filled by a method discussed in Brunetti et
al., 2004 and monthly series were calculated.
HOMOGENEITY TESTING AND RECORD ADJUSTING
Monthly records were tested for homogeneity by means of a procedure that rejects the apriori-existence of a homogeneous reference series. The resulting break signals were then
collected in a decision matrix and checked against metadata (Böhm et al., 2001).
Maps of the three rotated EOFs for Mean Temperature
STATION CLUSTERING AND REGIONAL AVERAGE SERIES
Temperature and precipitation records were clustered into homogeneous regions by
means of a Principal Component Analysis (PCA). Station records of the same area were
then averaged in order to obtain regional average series. Some results and examples are
given in the figures.
Maps of the six rotated EOFs for Precipitation
TREND ANALYSIS
Map of the Precipitation sub-regions based on PCA
Trends slopes were calculated by means of least square linear fitting, and their significance
levels were estimated by progressive non-parametric Mann–Kendall test (Sneyers, 1990).
Some examples of the results are given in the figures.
Full details on the methods and trend results are discussed in Brunetti et al., 2005.
1.5
1.5
Yearly maximum temperature for 3 Italian regions
K
A gridded version of the new data-sets was produced. The grid has one-degree resolution,
both in latitude and in longitude, and was realised with a Gaussian weighting function.
1.0
1.0
0.5
0.5
0.0
0.0
K
GRIDDING
Yearly minimum temperature for 3 Italian regions
-0.5
Map of the Temperature sub-regions based on PCA
RESULTS AND CONCLUSIONS
-0.5
-1.0
-1.0
NATIONAL AVERAGE
NATIONAL AVERAGE
PO PLAIN REGION M EAN SERIES
PO PLAIN REGION M EAN SERIES
ALPINE REGION M EAN SERIES
-1.5
RECORD ADJUSTMENTS
Gaussian low pass filtered curves (std deviation: 3 years)
1865
1880
1895
1910
1925
ALPINE REGION M EAN SERIES
PENINSULAR ITALY M EAN SERIES
1940
Year
1955
Our research highlighted that most Italian original temperature series are systematically biased by non-climatic noise. So, using the original data in estimating
long-term temperature evolution produces biased results. This result is in agreement with the findings of Böhm et al., 2001 and Begert et al., 2005.
1970
1985
-1.5
2000
Gaussian low pass filtered curves (std deviation: 3 years)
1865
1880
1895
1910
1925
PENINSULAR ITALY M EAN SERIES
1940
Year
1955
1970
1985
1.5
Yearly mean temperature for 3 Italian regions
1.0
STATION CLUSTERING
0.5
K
Principal Component Analysis allows some climatic regions to be identified.
In particular for minimum, mean and maximum temperatures, 3 EOFs, representing Po Plain (PP), Alpine (AL) and Peninsular Italy (PI) regions were extracted.
6 EOFs were extracted for precipitation. The identified regions are: North-Western Italy (NW), the Northern part of North-Eastern Italy (NEN), the Southern part
of North-Eastern Italy (NES), the Central Italy (CE), the South-Eastern Italy (SE) and the Southern Italy (SO).
0.0
-0.5
-1.0
NATIONAL AVERAGE
PO PLAIN REGION M EAN SERIES
TREND ANALYSIS
ALPINE REGION M EAN SERIES
-1.5
Quite a uniform temperature trend was observed in the different regions, with a trend of 1 K per century all over Italy on a yearly basis. The trend is
generally higher for minimum temperature than for maximum temperature for all seasons and the year.
Precipitation trend analysis showed a decreasing tendency, though such decreases are very low and rarely significant. Considering the average all over Italy
there is a 5% decrease per century in the annual precipitation amount, mainly due to the spring season (-9% per century).
A progressive trend analysis revealed that, both for temperature and precipitation, the significance and the slope of the trends strictly depend on the selected
period.
Temperature and precipitation trends obtained by analysing the new data-set show significant differences from the results presented in Maugeri and Nanni,
1998, in Buffoni et al., 1999 and in Brunetti et al., 2000, the old trend assessments being characterised by a lower temperature increase (around 0.5 K per
century) and a greater precipitation decrease (around 10% per century).
Tables show the estimated trends of minimum, mean, maximum temperature and precipitation in 1865-2003 period according to the new data-set. Bold
values indicate trends with significance level higher than 99%. For precipitation, the values are expressed in percentage per century relative to the mean of
the standard period 1961-1990.
Gaussian low pass filtered curves (std deviation: 3 years)
1865
1880
1895
1910
1925
PENINSULAR ITALY M EAN SERIES
1940
Year
1955
1970
1985
2000
Average regional series for mean, maximum, and minimum yearly temperature series (expressed in anomalies respect to
the mean of the standard period 1961-1990). The series were filtered with an 11-year window 3-year Gaussian low-pass
filter.
1.35
NATIONAL AVERAGE
NORTH WEST ITALY M EAN SERIES
NORTH-EAST-NORTH ITALY M EAN SERIES
NORTH-EAST-SOUTH ITALY M EAN SERIES
CENTER ITALY M EAN SERIES
SOUTH-EAST ITALY M EAN SERIES
SOUTH ITALY M EAN SERIES
1.30
1.25
1.20
Yearly total precipitation for 6 Italian regions
1.15
1.10
1.05
1.00
0.95
0.90
MINIMUM
MEAN
MAXIMUM
PRECIPITATION
0.85
North-Western Italy
-
0.80
Northern part of North-Eastern Italy
-
Southern part of North-Eastern Italy
-
Po Plain
TEMPERATURE
0.9 ± 0.1 (ºC / 100 y)
TEMPERATURE
1.0 ± 0.1 (ºC / 100 y)
TEMPERATURE
1.1 ± 0.1 (ºC / 100 y)
Alps
1.2 ± 0.1 (ºC / 100 y)
1.0 ± 0.1 (ºC / 100 y)
0.8 ± 0.1 (ºC / 100 y)
Peninsular Italy
1.3 ± 0.1 (ºC / 100 y)
1.0 ± 0.1 (ºC / 100 y)
0.7 ± 0.1 (ºC / 100 y)
Southern Italy
ITALY
1.1 ± 0.1 (ºC / 100 y)
1.0 ± 0.1 (ºC / 100 y)
0.9 ± 0.1 (ºC / 100 y)
ITALY
Gaussian low pass filtered curves (std deviation: 3 years)
Regional averages are calculated from at least 5 station records
The National average is the mean of the regional curves
1750
Central Italy
-(10 ± 3)
South-Eastern Italy
-(8 ± 5)
1775
1800
1825
1850
1875
Year
1900
1925
1950
1975
2000
Average regional yearly total precipitation series (expressed in ratios respect to the mean of the
standard period 1961-1990). The series were filtered with an 11-year window 3-year Gaussian lowpass filter.
+
-(5 ± 3)
40
Progressive slopes (%/100 y) of annual total precipitation series
Precipitation trends over the 1865-2003 period. The values are
expressed in percentage per century relative to the mean of the
standard period 1961-1990. Only values with significance level greater
than 90% are indicated; for lower values of significance only the sign of
the trend is indicated. Bold values indicate trends with significance level
higher than 99%.
Trends of mean, maximum and minimum temperatures over the 1865-2003 period. The
values are expressed in K per century. Bold values indicate trends with significance
level higher than 99%.
20
5
5
Progressive slopes (K/100 years) of
annual maximum temperature series
5
Progressive slopes (K/100 years) of
annual minimum temperature series
Progressive slopes (K/100 years) of
annual mean temperature series
3
3
3
2
1
1895
1910
1925
Year
1940
-40
2
1
1880
-20
K/100y
4
K/100y
4
K/100y
4
2
%/100y
0
0
1865
2000
1
NATIONAL AVER AGE
NATIONAL AVER AGE
NATIONAL AVER AGE
P O P LAIN R EGION M EAN S ER IES
P O P LAIN R EGION M EAN S ER IES
P O P LAIN R EGION M EAN S ER IES
ALP INE R EGION M EAN S ER IES
ALP INE R EGION M EAN S ER IES
ALP INE R EGION M EAN S ER IES
P ENINS ULAR ITALY M EAN S ER IES
P ENINS ULAR ITALY M EAN S ER IES
P ENINS ULAR ITALY M EAN S ER IES
1955
1970
0
1865
1880
1895
1910
1925
Year
1940
1955
1970
0
1865
1880
1895
1910
1925
Year
1940
1955
1970
Progressive slopes (K/100 years) of annual temperature records. Curves refer to the series starting from 1865 + i (i = 0, ….., 100) and ending in 2003. The slopes of the series are displayed in correspondence
of their starting year. The upper limit for i has been selected in order to perform trend analysis over at least 40-year periods. Bold parts indicate slopes significant at a 95% level.
-60
1750
1800
1850
1900
1950
Year
Progressive trend analysis for annual precipitation records. Each point of a curve represents the regression
coefficient (expressed in percentage per century relative to the mean of the standard period 1961-1990) calculated
for the series beginning from the year in which it is located and ending with the last one (2003). Thick portions of the
curves indicate trends with s.l.>95%.
REFERENCES
REFERENCES
Begert M., Schlegel T., Kirchhofer W., 2005. Homogeneous temperature and precipitation series of Switzerland from 1864 to 2000. Int. J. Climatol., 25, 65-80.
Begert
M.,Auer
Schlegel
T., Kirchhofer
W., 2005.
Homogeneous
temperature
and precipitation
series of Switzerland
1864
to 2000.
Int.
J. Climatol.
65-80. instrumental time series. Int. J. Climatol., 21, 1779-1801.
Böhm R.,
I., Brunetti
M., Maugeri
M., Nanni
T., Schöner
W., 2001.
Regional Temperature
Variabilityfrom
in the
European
Alps
1760-1998
from, 25,
homogenised
Böhm R., Auer I., Brunetti M., Maugeri M., Nanni T., Schöner W., 2001. Regional Temperature Variability in the European Alps 1760-1998 from homogenised instrumental time series. Int. J. Climatol., 21, 1779-1801.
Brunetti M., Buffoni L., Maugeri M., Nanni T., 2000. Trends of minimum and maximum daily temperatures in Italy from 1865 to 1996. Theor. Appl. Climatol., 66, 49-60.
Brunetti M., Buffoni L., Maugeri M., Nanni T., 2000. Trends of minimum and maximum daily temperatures in Italy from 1865 to 1996. Theor. Appl. Climatol., 66, 49-60.
Brunetti
Brunetti M.,
M., Maugeri
Maugeri M.,
M., Monti
Monti F.,
F., Nanni
Nanni T.,
T., 2004.
2004. Changes
Changes in
in daily
daily precipitation
precipitation frequency
frequency and
and distribution
distribution in
in Italy
Italy over
over the
the last
last 120
120 years.
years. J.
J. Geophys.
Geophys. Res.
Res.,, 109,
109, D05102,
D05102, doi:10.1029/2003JD004296.
doi:10.1029/2003JD004296.
Brunetti
Brunetti M.,
M., Maugeri
Maugeri M.,
M., Monti
Monti F.,
F., Nanni
Nanni T.,
T., 2005.
2005. Temperature
Temperature and
and precipitation
precipitation variability
variability in
in Italy
Italy in
in the
the last
last two
two centuries
centuries from
from homogenized
homogenized time
time series.
series. Int.
Int. J.
J. Climatol.
Climatol.,, 25,
25, (in
(in press).
press).
Buffoni L., Maugeri M., Nanni T., 1999. Precipitation in Italy from 1833 to 1996. Theor. Appl. Climatol., 63, 33-40.
Maugeri M., Nanni T., 1998. Surface air temperature variations in Italy: recent trends and an update to 1993. Theor. Appl. Climatol., 61, 191-196.