APPLICATIONIN CLIMATOLOGY 2:

LONG-TERM TRENDS IN PERSISTENCE

APPLICATIONIN CLIMATOLOGY 2:

LONG-TERM TRENDS IN PERSISTENCE

Radan HUTH,

Monika CAHYNOVÁ,

Jan KYSELÝ

Radan HUTH,

Monika CAHYNOVÁ,

Jan KYSELÝ

Hess&Brezowsky groups of types

dashed: lifetime (persistence)

smoothed

DJF

Hess&Brezowsky groups of types

dashed: lifetime (persistence)

smoothed

JJA

Hess&Brezowsky: groups of types with cyclonic / anticyclonic character over

central Europe

dashed: lifetime (persistence)

smoothed

Hess&Brezowsky: all types

lifetime (persistence)

Application in climatology 3: Links between

circulation changes and climatic trends

in Europe

Application in climatology 3: Links between

circulation changes and climatic trends

in Europe

OutlineOutline

we want to assess the magnitude of climatic trends over Europe in 1961-2000 that can be linked to changing frequency of circulation types (as opposed to changing climatic properties of circulation types)

data– 29 stations from the ECA&D project, daily Tmax, Tmin, precipitation

– 8 objective catalogues from cat.1.2 (CKMEANS, GWT, Litynski, LUND, P27, PETISCO, SANDRA, TPCA), each in 3 variants with 9, 18, 27 CTs

– all COST733 domains except for D03 – lack of stations

methods– seasonal climatic trends from station data

– proportion of climatic trends linked to circulation changes

we want to assess the magnitude of climatic trends over Europe in 1961-2000 that can be linked to changing frequency of circulation types (as opposed to changing climatic properties of circulation types)

data– 29 stations from the ECA&D project, daily Tmax, Tmin, precipitation

– 8 objective catalogues from cat.1.2 (CKMEANS, GWT, Litynski, LUND, P27, PETISCO, SANDRA, TPCA), each in 3 variants with 9, 18, 27 CTs

– all COST733 domains except for D03 – lack of stations

methods– seasonal climatic trends from station data

– proportion of climatic trends linked to circulation changes

StationsStations

Trends in the frequency of CTsTrends in the frequency of CTs

0

20

40

60

80

100

%

D00 D01 D02 D03 D04 D05 D06 D07 D08 D09 D10 D11

+ MAM + JJA + SON + DJF

Percentage of days occupied by CTs with trends in the seasonal frequency significant at the 95% level in 1961-2000

Trends in the frequency of CTsTrends in the frequency of CTs

Magnitude of significant trends in frequency of CTs in GWTC10 (days per season in 1961-2000)

W SW NW C A N NE E SE SD00 -5D01 -8D02 6 -3D03 -5D04D05D06D07D08D09 -8 5 3D10D11D00D01 -4D02D03D04D05D06 2D07 -2D08D09 -7D10 -2D11

spri

ngsu

mm

er

W SW NW C A N NE E SE SD00 9D01D02D03 3 4D04D05D06D07 3D08 3D09D10D11D00 16 -8 -3D01 6D02 7D03D04 13 13 -7 -7D05D06 9 3 -8D07 20 -6 -7D08 16 -3 -10 -8D09D10 5 -4 2 -6D11 14 -7

autu

mn

win

ter

Results – seasonal climatic trendsResults – seasonal climatic trendstrend significant at the 95% level

Ratio of “hypothetical” (circulation-induced) and observed long-term seasonal trends.

The “hypothetical” trend is calculated from a daily series, constructed by assigning the long-term monthly mean of the given variable under the specific circulation type to each day.

See e.g. Huth (2001).

Method to attribute climatic trends to changes in frequency of circulation

types

Method to attribute climatic trends to changes in frequency of circulation

types

-8

-6

-4

-2

0

2

4

6

8

2 2 2 2 1 2 1 1 1 2 8 5 1 5 5 8 8 9 10 10 10 10 9 8 8 8 2 2 1 1 1

de

gre

es

Ce

lsiu

s

observed Tmax

January average Tmax under each CT

Ratio of circulation-induced (“hypothetical”) and observed trends 1961-2000 at stations where the observed trend is significant at

the 95% levelResults of 24 classifications on D00 and small domains

Averages of individual stations where observed trends are significant at the 95% level

Ratio of circulation-induced (“hypothetical”) and observed trends 1961-2000

Comparison of individual classifications

TX

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

CK

ME

AN

SC

09C

KM

EA

NS

C18

CK

ME

AN

SC

27G

WT

C10

GW

TC

18G

WT

C26

LITA

DV

ELI

TC

18LI

TT

CLU

ND

C09

LUN

DC

18LU

ND

C27

P27

C08

P27

C16

P27

C27

PE

TIS

CO

C09

PE

TIS

CO

C18

PE

TIS

CO

C27

SA

ND

RA

C09

SA

ND

RA

C18

SA

ND

RA

C27

TP

CA

C09

TP

CA

C18

TP

CA

C27

rati

o

MAM-Dxx JJA-Dxx SON-Dxx DJF-Dxx

MAM-D00 JJA-D00 SON-D00 DJF-D00TN

-0.3

-0.2

-0.1

0

0.1

0.2

0.3

0.4

0.5

0.6

CK

ME

AN

SC

09C

KM

EA

NS

C18

CK

ME

AN

SC

27G

WT

C10

GW

TC

18G

WT

C26

LITA

DV

ELI

TC

18LI

TT

CLU

ND

C09

LUN

DC

18LU

ND

C27

P27

C08

P27

C16

P27

C27

PE

TIS

CO

C09

PE

TIS

CO

C18

PE

TIS

CO

C27

SA

ND

RA

C09

SA

ND

RA

C18

SA

ND

RA

C27

TP

CA

C09

TP

CA

C18

TP

CA

C27

rati

o

ConclusionsConclusions

Significant trends in the frequency of CTs occur mostly in winter in domains 00 and 04 through 11, and also in summer in the Mediterranean.

Climatic trends can be only partly explained by the changing frequency of CTs, the link being the strongest in winter. In the other seasons, within-type climatic trends are responsible for a major part of the observed trends.

Classifications in the small domains are usually more tightly connected with climatic trends than those in D00, except for the northernmost stations.

There are large differences between results obtained with individual classifications – therefore all studies using just a limited number of them should be taken with a grain of salt.

Application in climatology 4: Analysis of climate

model outputs

Application in climatology 4: Analysis of climate

model outputs

How to compare circulation types between two climates?How to compare circulation

types between two climates?

Isn’t it nonsense? We have just one climate...

Comparisons between– real climate and simulated present climate

model validation– simulated present and perturbed (typically

future) climate climate change response– real climate in two distinct periods (e.g.,

current vs. little ice age)

Isn’t it nonsense? We have just one climate...

Comparisons between– real climate and simulated present climate

model validation– simulated present and perturbed (typically

future) climate climate change response– real climate in two distinct periods (e.g.,

current vs. little ice age)

OBS CTR

“INSTRINSIC” TYPES

OBSERVED TYPES

projected onto CONTROL

OBS OBS CTR

BUT: isn’t it an artefact of the

projection?

projection in the opposite “direction”:

CONTROL TYPES

projected onto OBSERVED

CTRCTR OBS

How to compare circulation types between two climates?How to compare circulation

types between two climates?

(at least) four possible approaches

1. Find circulation types in each climate separately+ you may get truly dominant types in both

datasets (if you are lucky...)– no clear structure in data types are to a

certain extent random comparison may be misleading

(at least) four possible approaches

1. Find circulation types in each climate separately+ you may get truly dominant types in both

datasets (if you are lucky...)– no clear structure in data types are to a

certain extent random comparison may be misleading

How to compare circulation types between two climates?How to compare circulation

types between two climates?

2. Use types defined a priori, independently of the datasets

objectivized catalogues

types defined on a short(er) period

+ easy and fair comparison– may not reflect real structure in either

dataset

2. Use types defined a priori, independently of the datasets

objectivized catalogues

types defined on a short(er) period

+ easy and fair comparison– may not reflect real structure in either

dataset

How to compare circulation types between two climates?How to compare circulation

types between two climates?

3. Concatenation of two datasets, “joint” classification performed simultaneously for the two climates

(typically used with SOMs)+ good compromise: types are likely to be

close to ‘real’ types in both datasets

3. Concatenation of two datasets, “joint” classification performed simultaneously for the two climates

(typically used with SOMs)+ good compromise: types are likely to be

close to ‘real’ types in both datasets

How to compare circulation types between two climates?How to compare circulation

types between two climates?

4. Projection from one climate to the other and vice versa

+ wrong conclusions are eliminated

4. Projection from one climate to the other and vice versa

+ wrong conclusions are eliminated

Where to find it? ReferencesWhere to find it? References

Huth R. et al., 2008: Classifications of atmospheric circulation patterns: Recent advances and applications. Ann. N. Y. Acad. Sci., 1146, 105-152.

ad 1) (heat waves)Kyselý J., Huth R., 2008: Adv. Geosci., 14, 243-249.

ad 2) (trends in persistence)Kyselý J., Huth R., 2006: Theor. Appl. Climatol., 85, 19-36. Cahynová M., Huth R., 2009: Tellus A, 61, 407-416.

ad 3) (climate change vs. circulation)Huth R., 2001: Int. J. Climatol., 21, 135-153. Cahynová M., Huth R., 2009: Theor. Appl. Climatol., 96, 57-68.

ad 4) (analysis of GCM outputs)Huth R., 1997: J. Climate, 10, 1545-1561. Huth R., 2000: Theor. Appl. Climatol., 67, 1-18.

Huth R. et al., 2008: Classifications of atmospheric circulation patterns: Recent advances and applications. Ann. N. Y. Acad. Sci., 1146, 105-152.

ad 1) (heat waves)Kyselý J., Huth R., 2008: Adv. Geosci., 14, 243-249.

ad 2) (trends in persistence)Kyselý J., Huth R., 2006: Theor. Appl. Climatol., 85, 19-36. Cahynová M., Huth R., 2009: Tellus A, 61, 407-416.

ad 3) (climate change vs. circulation)Huth R., 2001: Int. J. Climatol., 21, 135-153. Cahynová M., Huth R., 2009: Theor. Appl. Climatol., 96, 57-68.

ad 4) (analysis of GCM outputs)Huth R., 1997: J. Climate, 10, 1545-1561. Huth R., 2000: Theor. Appl. Climatol., 67, 1-18.