Meteorologische Zeitschrift Vol15 No 3 259-263 (June 2006)ccopy by Gebruumlder Borntraeger 2006 Article

World Map of the Koumlppen-Geiger climate classificationupdated

MARKUS KOTTEK1 JUumlRGENGRIESER2 CHRISTOPHBECK2 BRUNO RUDOLF2 and FRANZ RUBELlowast1

1Biometeorology Group University of Veterinary Medicine Vienna Vienna Austria2Global Precipitation Climatology Centre Deutscher Wetterdienst Offenbach Germany

(Manuscript received December 19 2005 in revised form February 28 2006 accepted April 10 2006)

AbstractThe most frequently used climate classification map is that of Wladimir Koumlppen presented in its latest version1961 by Rudolf Geiger A huge number of climate studies and subsequent publications adopted this or aformer release of the Koumlppen-Geiger map While the climate classification concept has been widely appliedto a broad range of topics in climate and climate change research as well as in physical geography hydrologyagriculture biology and educational aspects a well-documented update of the world climate classificationmap is still missing Based on recent data sets from the Climatic Research Unit (CRU) of the University ofEast Anglia and the Global Precipitation Climatology Centre (GPCC) at the German Weather Service wepresent here a new digital Koumlppen-Geiger world map on climate classification valid for the second half ofthe 20th century

ZusammenfassungDie am haumlufigsten verwendete Klimaklassifikationskarte istjene von Wladimir Koumlppen die in der letztenAuflage von Rudolf Geiger aus dem Jahr 1961 vorliegt Seitherbildeten viele Klimabuumlcher und Fachartikeldiese oder eine fruumlhere Ausgabe der Koumlppen-Geiger Karte abObwohl das Schema der Klimaklassifikationin vielen Forschungsgebieten wie Klima und Klimaaumlnderung aber auch physikalische Geographie Hydrolo-gie Landwirtschaftsforschung Biologie und Ausbildung zum Einsatz kommt fehlt bis heute eine gut doku-mentierte Aktualisierung der Koumlppen-Geiger Klimakarte Basierend auf neuesten Datensaumltzen des ClimaticResearch Unit (CRU) der Universitaumlt von East Anglia und des Weltzentrums fuumlr Niederschlagsklimatologie(WZN) am Deutschen Wetterdienst praumlsentieren wir hier eineneue digitale Koumlppen-Geiger Weltkarte fuumlr diezweite Haumllfte des 20 Jahrhunderts

1 Introduction

The first quantitative classification of world climateswas presented by the German scientist Wladimir Koumlp-pen (1846ndash1940) in 1900 it has been available asworld map updated 1954 and 1961 by Rudolf Geiger(1894ndash1981) Many of the early German publications(KOumlPPEN 1900 GEIGER1954 1957) from this area arenot easily accessible today here we refer to the compre-hensive summaries on this topic given by eg HANTEL

(1989) or ESSENWANGER(2001)Koumlppen was trained as a plant physiologist and re-

alised that plants are indicators for many climatic el-ements His effective classification was constructed onthe basis of five vegetation groups determined by theFrench botanist De Candolle referring to the climatezones of the ancient Greeks (SANDERSON 1999) Thefive vegetation groups of Koumlppen distinguish betweenplants of the equatorial zone (A) the arid zone (B) thewarm temperate zone (C) the snow zone (D) and the po-lar zone (E) A second letter in the classification consid-

lowastCorresponding author Franz Rubel Biometeorology Group De-partment of Natural Sciences University of Veterinary Medicine Vi-enna 1210 Vienna Austria e-mail franzrubelvu-wienacat

ers the precipitation (eg Df for snow and fully humid)a third letter the air temperature (eg Dfc for snow fullyhumid with cool summer)

Although various authors published enhanced Koumlp-pen classifications or developed new classifications theclimate classification originally developed by Koumlppen(here referred to as Koumlppen-Geiger classification) is stillthe most frequently used climate classification Manytextbooks on climatology reproduce a world map ofKoumlppen-Geiger climate classes due to the lack of recentmaps mostly a copy of one of the historical hand-drawnmaps (eg KRAUS 2004) In order to close this gapwe present a digital world map of the Koumlppen-Geigerclimate classification calculated from up-to-date globaltemperature and precipitation data sets

The importance of an updated digital map may berecognized by looking at global and regional studies thatuse the Koumlppen-Geiger climate classification Represen-tative for hydrological studies PEEL et al (2001) iden-tified and explained the continental-scale variability inannual runoff by applying Koumlppenrsquos climate classifica-tion Applications to climate modelling have been pre-sented for example by LOHMANN et al (1993) to val-idate general circulation model control runs of present

DOI 1011270941-294820060130

0941-294820060130 $ 225

ccopy Gebruumlder Borntraeger Berlin Stuttgart 2006

260 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated Meteorol Z15 2006

Table 1 Key to calculate the climate formula of Koumlppen and Geiger for the main climatesand subsequent precipitation conditions the first

two letters of the classification Note that for the polar climates (E) no precipitation differentiations are given only temperature conditions

are defined This key implies that the polar climates (E) have to be determined first followed by the arid climates (B) and subsequent

differentiations into the equatorial climates (A) and the warm temperate and snow climates (C) and (D) respectively The criteria are

explained in the text

Type Description Criterion

A Equatorial climates Tmin ge +18 CAf Equatorial rainforest fully humid Pmin ge 60 mmAm Equatorial monsoon Pannge 25(100minusPmin)As Equatorial savannah with dry summer Pmin lt 60 mm in summerAw Equatorial savannah with dry winter Pmin lt 60 mm in winter

B Arid climates Pannlt 10 PthBS Steppe climate Panngt 5 PthBW Desert climate Pannle 5 Pth

C Warm temperate climates minus3 C lt Tmin lt +18 CCs Warm temperate climate with dry summer Psmin lt Pwmin Pwmax gt 3 Psmin and Psmin lt 40 mmCw Warm temperate climate with dry winter Pwmin lt Psmin and Psmaxgt 10 PwminCf Warm temperate climate fully humid neither Cs nor Cw

D Snow climates Tmin leminus3 CDs Snow climate with dry summer Psmin lt Pwmin Pwmax gt 3 Psmin and Psmin lt 40 mmDw Snow climate with dry winter Pwmin lt Psmin and Psmaxgt 10 PwminDf Snow climate fully humid neither Ds nor Dw

E Polar climates Tmax lt +10 CET Tundra climate 0C le Tmax lt +10 CEF Frost climate Tmax lt 0 C

climate as well as greenhouse gas warming simulationsKLEIDON et al (2000) investigated the maximum possi-ble influence of vegetation on the global climate by con-ducting climate model simulations Both LOHMANN etal (1993) and KLEIDON et al (2000) applied the Koumlp-pen classification to model simulations to illustrate thedifferences in simulation results The updated Koumlppen-Geiger climates presented here will support future stud-ies similar to those discussed above

2 Data and method

Two global data sets of climate observations have beenselected to update the historical world map of theKoumlppen-Geiger climate classes Both are available on aregular 05 degree latitudelongitude grid with monthlyresolution The first data set is provided by the Cli-matic Research Unit (CRU) of the University of EastAnglia (MITCHELL and JONES 2005) and delivers gridsof monthly climate observations from meteorologicalstations comprising nine climate variables from whichonly temperature is used in this study The temperaturefields have been analysed from time-series observations

which are checked for inhomogeneities in the station-records by an automated method This data set coversthe global land areas excluding Antarctica It is publiclyavailable (wwwcruueaacuk) and will be referred to asCRU TS 21

The second data set (BECK et al 2005) is pro-vided by the Global Precipitation Climatology Centre(GPCC) located at the German Weather Service Thisnew gridded monthly precipitation data set covers theglobal land areas excluding Greenland and AntarcticaIt was developed on the basis of the most comprehen-sive data-base of monthly observed precipitation dataworld-wide built by the GPCC All observations in thisstation data base are subject to a multi-stage qualitycontrol to minimise the risk of generating temporal in-homogeneities in the gridded data due to varying sta-tion densities This dataset is referred to as VASClimOv111 and is also freely available for scientific purposes(httpgpccdwdde) Both CRU TS 21 and VASClimOv11 data cover the 50-year period 1951 to 2000 se-lected in this study for updating the Koumlppen-Geigermap

1Variability Analysis of Surface Climate Observations

260 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated Meteorol Z15 2006

Table 1 Key to calculate the climate formula of Koumlppen and Geiger for the main climatesand subsequent precipitation conditions the first

two letters of the classification Note that for the polar climates (E) no precipitation differentiations are given only temperature conditions

are defined This key implies that the polar climates (E) have to be determined first followed by the arid climates (B) and subsequent

differentiations into the equatorial climates (A) and the warm temperate and snow climates (C) and (D) respectively The criteria are

explained in the text

Type Description Criterion

A Equatorial climates Tmin ge +18 CAf Equatorial rainforest fully humid Pmin ge 60 mmAm Equatorial monsoon Pannge 25(100minusPmin)As Equatorial savannah with dry summer Pmin lt 60 mm in summerAw Equatorial savannah with dry winter Pmin lt 60 mm in winter

B Arid climates Pannlt 10 PthBS Steppe climate Panngt 5 PthBW Desert climate Pannle 5 Pth

C Warm temperate climates minus3 C lt Tmin lt +18 CCs Warm temperate climate with dry summer Psmin lt Pwmin Pwmax gt 3 Psmin and Psmin lt 40 mmCw Warm temperate climate with dry winter Pwmin lt Psmin and Psmaxgt 10 PwminCf Warm temperate climate fully humid neither Cs nor Cw

D Snow climates Tmin leminus3 CDs Snow climate with dry summer Psmin lt Pwmin Pwmax gt 3 Psmin and Psmin lt 40 mmDw Snow climate with dry winter Pwmin lt Psmin and Psmaxgt 10 PwminDf Snow climate fully humid neither Ds nor Dw

E Polar climates Tmax lt +10 CET Tundra climate 0C le Tmax lt +10 CEF Frost climate Tmax lt 0 C

climate as well as greenhouse gas warming simulationsKLEIDON et al (2000) investigated the maximum possi-ble influence of vegetation on the global climate by con-ducting climate model simulations Both LOHMANN etal (1993) and KLEIDON et al (2000) applied the Koumlp-pen classification to model simulations to illustrate thedifferences in simulation results The updated Koumlppen-Geiger climates presented here will support future stud-ies similar to those discussed above

2 Data and method

Two global data sets of climate observations have beenselected to update the historical world map of theKoumlppen-Geiger climate classes Both are available on aregular 05 degree latitudelongitude grid with monthlyresolution The first data set is provided by the Cli-matic Research Unit (CRU) of the University of EastAnglia (MITCHELL and JONES 2005) and delivers gridsof monthly climate observations from meteorologicalstations comprising nine climate variables from whichonly temperature is used in this study The temperaturefields have been analysed from time-series observations

which are checked for inhomogeneities in the station-records by an automated method This data set coversthe global land areas excluding Antarctica It is publiclyavailable (wwwcruueaacuk) and will be referred to asCRU TS 21

The second data set (BECK et al 2005) is pro-vided by the Global Precipitation Climatology Centre(GPCC) located at the German Weather Service Thisnew gridded monthly precipitation data set covers theglobal land areas excluding Greenland and AntarcticaIt was developed on the basis of the most comprehen-sive data-base of monthly observed precipitation dataworld-wide built by the GPCC All observations in thisstation data base are subject to a multi-stage qualitycontrol to minimise the risk of generating temporal in-homogeneities in the gridded data due to varying sta-tion densities This dataset is referred to as VASClimOv111 and is also freely available for scientific purposes(httpgpccdwdde) Both CRU TS 21 and VASClimOv11 data cover the 50-year period 1951 to 2000 se-lected in this study for updating the Koumlppen-Geigermap

1Variability Analysis of Surface Climate Observations

Meteorol Z15 2006 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated 261

minus1

60

minus1

40

minus1

20

minus1

00

minus8

0minus

60

minus4

0minus

20

02

04

06

08

010

012

014

016

018

0

minus1

60

minus1

40

minus1

20

minus1

00

minus8

0minus

60

minus4

0minus

20

02

04

06

08

010

012

014

016

018

0

minus80

minus70

minus60

minus5

0

minus40

minus30minus20

minus100

10

20

30

40

50

60

70

80

minus9

0

minus80

minus70

minus60

minus50

minus40

minus3

0

minus20

minus10

0102

030

40

50

60

70

80

90

Af

Am

As

Aw

BW

kB

Wh

BS

kB

Sh

Cfa

Cfb

Cfc

Csa

Csb

Csc

Cw

a

Cw

bC

wc

Dfa

Dfb

Dfc

Dfd

Dsa

Dsb

Dsc

Dsd

Dw

aD

wb

Dw

cD

wd

EF

ET

Worl

d M

ap

of

Koumlp

pen

minusG

eiger

Cli

mate

Cla

ssif

icati

on

up

da

ted

wit

h C

RU

TS

21

tem

per

atu

re a

nd

VA

SC

lim

O v

11

pre

cip

ita

tion

da

ta 1

95

1 t

o 2

000

Main

cli

mate

s

A

equ

ato

rial

B

ari

d

C

wa

rm t

emp

era

te

D

sno

w

E

pola

r

Pre

cip

itati

on

W

des

ert

S

step

pe

f f

ull

y h

um

id

s s

um

mer

dry

w

win

ter

dry

m

mon

soo

na

l

Tem

per

atu

re

a

ho

t su

mm

er

b

wa

rm s

um

mer

c c

ool

sum

mer

d

extr

emel

y c

on

tin

enta

l

h

ho

t a

rid

k

cold

ari

d

F

pola

r fr

ost

T

pola

r tu

nd

ra

Res

olu

tio

n

05

deg

lat

lon

Figure 1 World Map of Koumlppen-Geiger climate classification updated with mean monthlyCRU TS 21 temperature and VASClimO v11

precipitation data for the period 1951 to 2000 on a regular 05 degree latitudelongitude grid

262 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated Meteorol Z15 2006

Table 2 Key to calculate the third letter temperature classification (h) and (k) for the arid climates (B) and (a) to (d) for the warm temperate

and snow climates (C) and (D) Note that for type (b) warm summer a threshold temperature value of +10C has to occur for at least four

months The criteria are explained in the text

Type Description Criterion

h Hot steppe desert Tannge +18 Ck Cold steppe desert Tannlt +18 C

a Hot summer Tmaxge +22 Cb Warm summer not (a) and at least 4 Tmonge +10 Cc Cool summer and cold winter not (b) and Tmin gt minus38 Cd extremely continental like (c) but Tmin leminus38 C

Since various different sometimes just slightly mod-ified versions of Koumlppenrsquos climate classification havebeen published the calculation scheme for the Koumlppen-Geiger classes as applied here will now be briefly de-scribed (for more details see eg section 1342 ofHANTEL 1989 KRAUS 2004) This guarantees thereproducibility of the digital data set presented hereThe key to the main climates characterized by the firsttwo letters of the classification is described in Tab 1The annual mean near-surface (2 m) temperature is de-noted by Tannand the monthly mean temperatures of thewarmest and coldest months by Tmax and Tmin respec-tively Pann is the accumulated annual precipitation andPmin is the precipitation of the driest month AdditionallyPsmin Psmax Pwmin and Pwmax are defined as the lowestand highest monthly precipitation values for the summerand winter half-years on the hemisphere considered Alltemperatures are given inC monthly precipitations inmmmonth and Pann in mmyear

In addition to these temperature and precipitationvalues a dryness threshold Pth in mm is introduced forthe arid climates (B) which depends onTann the ab-solute measure of the annual mean temperature inCand on the annual cycle of precipitation

Pth =

2Tann if at least 23 of the annualprecipitation occurs in winter

2Tann+28 if at least 23 of the annualprecipitation occurs in summer

2Tann+14 otherwise(21)

The scheme how to determine the additional temper-ature conditions (third letter) for the arid climates (B) aswell as for the warm temperate and snow climates (C)and (D) respectively is given in Tab 2 where Tmon de-notes the mean monthly temperature inC

3 Results

Combining the three letters depicted in Tab 1 and Tab2 leads to at most 34 possible different climate classes

Three of these classes cannot occur by definition sincea warm temperate climate (C) needs a temperature ofthe coldest month Tmin above ndash3C while a third let-ter climate (d) extremely continental needs a temper-ature of the coldest month below ndash38C Therefore(Csd) (Cwd) and (Cfd) cannot be realised and 31 cli-mate classes remain Koumlppen and Geiger recognised thatnot all of the remaining types occur in a large arealamount and therefore not all of these types may be ofclimatological importance

Fig 1 shows a world map of the Koumlppen-Geiger cli-mate classification updated with mean monthly CRUTS 21 temperature and VASClimO v11 precipitationdata for the period 1951 to 2000 on a regular 05 de-gree latitudelongitude grid All 31 climate classes areillustrated with different colours although one of theseclasses (Dsd) does never occur in this map and someothers (Cfc Csc Cwc Dsa Dsb and Dsc) occur onlyin very small areas Having neither temperature nor pre-cipitation data available for Antarctica this region hasbeen set manually to the polar frost climate (EF) bythe use of a 05 land-sea-mask operationally applied atthe GPCC Also for Greenland no precipitation data areavailable However this has no influence on the classi-fication since temperature data strongly suggest that theclimate of Greenland is either polar tundra (ET) or polarfrost (EF) and is therefore independent of precipitation(Tab 1)

The resulting world map depicted in Fig 1 corre-sponds quite well with the historical hand-drawn mapsof the Koumlppen-Geiger climates but shows more regionaldetails due to the high spatial resolution of 05 degreeand provides the opportunity for further investigationsby applying the underlying digital data For examplestudies on depicting global climate change have beenperformed by the authors and will be published soon

4 Conclusion

SANDERSON(1999) stated in the closing sentence of herreview paper on climate classificationsModern atlasesand geography textbooks continue to use the 100-year

Meteorol Z15 2006 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated 263

old Koumlppen classification of climate and she askedIs it not time for modern atmospheric scientists to de-velop a new classification of world climatesWe be-lieve that the climate classification concept developed inthe first half of the 20th century by Koumlppen and Geigeris not likely to be discarded in the next future in factit still appears to meet the needs of todayrsquos climate sci-entists (ESSENWANGER 2001 KRAUS 2004) Updatedon the basis of recent (HANTEL 2005) and future highresolution climate data and applied to climate modelpredictions (eg LOHMANN et al 1993 KLEIDON etal 2000) the Koumlppen-Geiger classification might havea good chance to be applicable for another 100 years

The world map of the Koumlppen-Geiger climate classi-fication presented here as well as the underlying digitaldata are publicly available and distributed by the GlobalPrecipitation Climatology Centre (GPCC) at the Ger-man Weather Service (httpgpccdwdde) and the Uni-versity of Veterinary Medicine Vienna (httpkoeppen-geigervu-wienacat)

Acknowledgements

The German Climate Research Programme (DEKLIM)of the Federal Ministry of Education and Research andthe FP6 Integrated project GEOLAND (SIP3-CT-2003-502871) funded parts of this work

References

BECK C J GRIESER B RUDOLF 2005 A New MonthlyPrecipitation Climatology for the Global Land Areas forthe Period 1951 to 2000 ndash Climate status report 2004181ndash190 German Weather Service Offenbach GermanyReprint available at httpgpccdwdde

ESSENWANGER O M 2001 Classification of ClimatesWorld Survey of Climatology 1C General Climatology ndashElsevier Amsterdam 102 pp

GEIGER R 1954 Landolt-Boumlrnstein ndash Zahlenwerte undFunktionen aus Physik Chemie Astronomie Geophysikund Technik alte Serie Vol3 Ch Klassifikation der Kli-mate nach W Koumlppen ndash Springer Berlin 603ndash607

mdash 1961 Uumlberarbeitete Neuausgabe von Geiger R Koumlppen-Geiger Klima der Erde (Wandkarte 116 Mill) ndash Klett-Perthes Gotha

HANTEL M 1989 Climatology Series Landolt-Boumlrnstein ndashNumerical Data and Functional Relationships in Scienceand Technology New Series Vol4C2 Ch The presentglobal surface climate ndash Springer Berlin 117ndash474

mdash (Ed) 2005 Observed Global Climate Series Landolt-Boumlrnstein ndash Numerical Data and Functional Relationshipsin Science and Technology New Series Vol6A ndash SpringerBerlin

KLEIDON A K FRAEDRICH M HEIMANN 2000 ndash Agreen planet versus a desert world Estimating the maxi-mum effect of vegetation on the land surface climate ndash Cli-mate Change44 471ndash493

KOumlPPEN W 1900 ndash Versuch einer Klassifikation der Kli-mate vorzugsweise nach ihren Beziehungen zur Pflanzen-welt ndash Geogr Zeitschr6 593ndash611 657ndash679

KRAUS H 2004 Die Atmosphaumlre der Erde Eine Einfuumlhrungin die Meteorologie ndash Springer Berlin 422 pp

LOHMANN U R SAUSEN L BENGTSSONU CUBASCH J PERLWITZ E ROECKNER 1993The Koumlppen climate classification as a diagnostic tool forgeneral circulation models ndash Climate Res3 177ndash193

M ITCHELL T D P D JONES 2005 An improved methodof constructing a database of monthly climate observationsand associated high-resolution grids ndash Int J Climatol25693ndash712

PEEL M C T A MCMAHON B L FINLAYSON F G R WATSON 2001 Identification and explanation ofcontinental differences in the variability of annual runoff ndashJ Hydrol250 224ndash240

SANDERSON M 1999 The classification of climates fromPythagoras to Koeppen ndash Bull Amer Meteor Soc80669ndash673

262 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated Meteorol Z15 2006

Table 2 Key to calculate the third letter temperature classification (h) and (k) for the arid climates (B) and (a) to (d) for the warm temperate

and snow climates (C) and (D) Note that for type (b) warm summer a threshold temperature value of +10C has to occur for at least four

months The criteria are explained in the text

Type Description Criterion

h Hot steppe desert Tannge +18 Ck Cold steppe desert Tannlt +18 C

a Hot summer Tmaxge +22 Cb Warm summer not (a) and at least 4 Tmonge +10 Cc Cool summer and cold winter not (b) and Tmin gt minus38 Cd extremely continental like (c) but Tmin leminus38 C

Since various different sometimes just slightly mod-ified versions of Koumlppenrsquos climate classification havebeen published the calculation scheme for the Koumlppen-Geiger classes as applied here will now be briefly de-scribed (for more details see eg section 1342 ofHANTEL 1989 KRAUS 2004) This guarantees thereproducibility of the digital data set presented hereThe key to the main climates characterized by the firsttwo letters of the classification is described in Tab 1The annual mean near-surface (2 m) temperature is de-noted by Tannand the monthly mean temperatures of thewarmest and coldest months by Tmax and Tmin respec-tively Pann is the accumulated annual precipitation andPmin is the precipitation of the driest month AdditionallyPsmin Psmax Pwmin and Pwmax are defined as the lowestand highest monthly precipitation values for the summerand winter half-years on the hemisphere considered Alltemperatures are given inC monthly precipitations inmmmonth and Pann in mmyear

In addition to these temperature and precipitationvalues a dryness threshold Pth in mm is introduced forthe arid climates (B) which depends onTann the ab-solute measure of the annual mean temperature inCand on the annual cycle of precipitation

Pth =

2Tann if at least 23 of the annualprecipitation occurs in winter

2Tann+28 if at least 23 of the annualprecipitation occurs in summer

2Tann+14 otherwise(21)

The scheme how to determine the additional temper-ature conditions (third letter) for the arid climates (B) aswell as for the warm temperate and snow climates (C)and (D) respectively is given in Tab 2 where Tmon de-notes the mean monthly temperature inC

3 Results

Combining the three letters depicted in Tab 1 and Tab2 leads to at most 34 possible different climate classes

Three of these classes cannot occur by definition sincea warm temperate climate (C) needs a temperature ofthe coldest month Tmin above ndash3C while a third let-ter climate (d) extremely continental needs a temper-ature of the coldest month below ndash38C Therefore(Csd) (Cwd) and (Cfd) cannot be realised and 31 cli-mate classes remain Koumlppen and Geiger recognised thatnot all of the remaining types occur in a large arealamount and therefore not all of these types may be ofclimatological importance

Fig 1 shows a world map of the Koumlppen-Geiger cli-mate classification updated with mean monthly CRUTS 21 temperature and VASClimO v11 precipitationdata for the period 1951 to 2000 on a regular 05 de-gree latitudelongitude grid All 31 climate classes areillustrated with different colours although one of theseclasses (Dsd) does never occur in this map and someothers (Cfc Csc Cwc Dsa Dsb and Dsc) occur onlyin very small areas Having neither temperature nor pre-cipitation data available for Antarctica this region hasbeen set manually to the polar frost climate (EF) bythe use of a 05 land-sea-mask operationally applied atthe GPCC Also for Greenland no precipitation data areavailable However this has no influence on the classi-fication since temperature data strongly suggest that theclimate of Greenland is either polar tundra (ET) or polarfrost (EF) and is therefore independent of precipitation(Tab 1)

The resulting world map depicted in Fig 1 corre-sponds quite well with the historical hand-drawn mapsof the Koumlppen-Geiger climates but shows more regionaldetails due to the high spatial resolution of 05 degreeand provides the opportunity for further investigationsby applying the underlying digital data For examplestudies on depicting global climate change have beenperformed by the authors and will be published soon

4 Conclusion

SANDERSON(1999) stated in the closing sentence of herreview paper on climate classificationsModern atlasesand geography textbooks continue to use the 100-year

Meteorol Z15 2006 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated 263

old Koumlppen classification of climate and she askedIs it not time for modern atmospheric scientists to de-velop a new classification of world climatesWe be-lieve that the climate classification concept developed inthe first half of the 20th century by Koumlppen and Geigeris not likely to be discarded in the next future in factit still appears to meet the needs of todayrsquos climate sci-entists (ESSENWANGER 2001 KRAUS 2004) Updatedon the basis of recent (HANTEL 2005) and future highresolution climate data and applied to climate modelpredictions (eg LOHMANN et al 1993 KLEIDON etal 2000) the Koumlppen-Geiger classification might havea good chance to be applicable for another 100 years

The world map of the Koumlppen-Geiger climate classi-fication presented here as well as the underlying digitaldata are publicly available and distributed by the GlobalPrecipitation Climatology Centre (GPCC) at the Ger-man Weather Service (httpgpccdwdde) and the Uni-versity of Veterinary Medicine Vienna (httpkoeppen-geigervu-wienacat)

Acknowledgements

The German Climate Research Programme (DEKLIM)of the Federal Ministry of Education and Research andthe FP6 Integrated project GEOLAND (SIP3-CT-2003-502871) funded parts of this work

References

BECK C J GRIESER B RUDOLF 2005 A New MonthlyPrecipitation Climatology for the Global Land Areas forthe Period 1951 to 2000 ndash Climate status report 2004181ndash190 German Weather Service Offenbach GermanyReprint available at httpgpccdwdde

ESSENWANGER O M 2001 Classification of ClimatesWorld Survey of Climatology 1C General Climatology ndashElsevier Amsterdam 102 pp

GEIGER R 1954 Landolt-Boumlrnstein ndash Zahlenwerte undFunktionen aus Physik Chemie Astronomie Geophysikund Technik alte Serie Vol3 Ch Klassifikation der Kli-mate nach W Koumlppen ndash Springer Berlin 603ndash607

mdash 1961 Uumlberarbeitete Neuausgabe von Geiger R Koumlppen-Geiger Klima der Erde (Wandkarte 116 Mill) ndash Klett-Perthes Gotha

HANTEL M 1989 Climatology Series Landolt-Boumlrnstein ndashNumerical Data and Functional Relationships in Scienceand Technology New Series Vol4C2 Ch The presentglobal surface climate ndash Springer Berlin 117ndash474

mdash (Ed) 2005 Observed Global Climate Series Landolt-Boumlrnstein ndash Numerical Data and Functional Relationshipsin Science and Technology New Series Vol6A ndash SpringerBerlin

KLEIDON A K FRAEDRICH M HEIMANN 2000 ndash Agreen planet versus a desert world Estimating the maxi-mum effect of vegetation on the land surface climate ndash Cli-mate Change44 471ndash493

KOumlPPEN W 1900 ndash Versuch einer Klassifikation der Kli-mate vorzugsweise nach ihren Beziehungen zur Pflanzen-welt ndash Geogr Zeitschr6 593ndash611 657ndash679

KRAUS H 2004 Die Atmosphaumlre der Erde Eine Einfuumlhrungin die Meteorologie ndash Springer Berlin 422 pp

LOHMANN U R SAUSEN L BENGTSSONU CUBASCH J PERLWITZ E ROECKNER 1993The Koumlppen climate classification as a diagnostic tool forgeneral circulation models ndash Climate Res3 177ndash193

M ITCHELL T D P D JONES 2005 An improved methodof constructing a database of monthly climate observationsand associated high-resolution grids ndash Int J Climatol25693ndash712

PEEL M C T A MCMAHON B L FINLAYSON F G R WATSON 2001 Identification and explanation ofcontinental differences in the variability of annual runoff ndashJ Hydrol250 224ndash240

SANDERSON M 1999 The classification of climates fromPythagoras to Koeppen ndash Bull Amer Meteor Soc80669ndash673

Meteorol Z15 2006 M Kottek et al World Map of the Koumlppen-Geiger climate classification updated 263

old Koumlppen classification of climate and she askedIs it not time for modern atmospheric scientists to de-velop a new classification of world climatesWe be-lieve that the climate classification concept developed inthe first half of the 20th century by Koumlppen and Geigeris not likely to be discarded in the next future in factit still appears to meet the needs of todayrsquos climate sci-entists (ESSENWANGER 2001 KRAUS 2004) Updatedon the basis of recent (HANTEL 2005) and future highresolution climate data and applied to climate modelpredictions (eg LOHMANN et al 1993 KLEIDON etal 2000) the Koumlppen-Geiger classification might havea good chance to be applicable for another 100 years

The world map of the Koumlppen-Geiger climate classi-fication presented here as well as the underlying digitaldata are publicly available and distributed by the GlobalPrecipitation Climatology Centre (GPCC) at the Ger-man Weather Service (httpgpccdwdde) and the Uni-versity of Veterinary Medicine Vienna (httpkoeppen-geigervu-wienacat)

Acknowledgements

The German Climate Research Programme (DEKLIM)of the Federal Ministry of Education and Research andthe FP6 Integrated project GEOLAND (SIP3-CT-2003-502871) funded parts of this work

References

BECK C J GRIESER B RUDOLF 2005 A New MonthlyPrecipitation Climatology for the Global Land Areas forthe Period 1951 to 2000 ndash Climate status report 2004181ndash190 German Weather Service Offenbach GermanyReprint available at httpgpccdwdde

ESSENWANGER O M 2001 Classification of ClimatesWorld Survey of Climatology 1C General Climatology ndashElsevier Amsterdam 102 pp

GEIGER R 1954 Landolt-Boumlrnstein ndash Zahlenwerte undFunktionen aus Physik Chemie Astronomie Geophysikund Technik alte Serie Vol3 Ch Klassifikation der Kli-mate nach W Koumlppen ndash Springer Berlin 603ndash607

mdash 1961 Uumlberarbeitete Neuausgabe von Geiger R Koumlppen-Geiger Klima der Erde (Wandkarte 116 Mill) ndash Klett-Perthes Gotha

HANTEL M 1989 Climatology Series Landolt-Boumlrnstein ndashNumerical Data and Functional Relationships in Scienceand Technology New Series Vol4C2 Ch The presentglobal surface climate ndash Springer Berlin 117ndash474

mdash (Ed) 2005 Observed Global Climate Series Landolt-Boumlrnstein ndash Numerical Data and Functional Relationshipsin Science and Technology New Series Vol6A ndash SpringerBerlin

KLEIDON A K FRAEDRICH M HEIMANN 2000 ndash Agreen planet versus a desert world Estimating the maxi-mum effect of vegetation on the land surface climate ndash Cli-mate Change44 471ndash493

KOumlPPEN W 1900 ndash Versuch einer Klassifikation der Kli-mate vorzugsweise nach ihren Beziehungen zur Pflanzen-welt ndash Geogr Zeitschr6 593ndash611 657ndash679

KRAUS H 2004 Die Atmosphaumlre der Erde Eine Einfuumlhrungin die Meteorologie ndash Springer Berlin 422 pp

LOHMANN U R SAUSEN L BENGTSSONU CUBASCH J PERLWITZ E ROECKNER 1993The Koumlppen climate classification as a diagnostic tool forgeneral circulation models ndash Climate Res3 177ndash193

M ITCHELL T D P D JONES 2005 An improved methodof constructing a database of monthly climate observationsand associated high-resolution grids ndash Int J Climatol25693ndash712

PEEL M C T A MCMAHON B L FINLAYSON F G R WATSON 2001 Identification and explanation ofcontinental differences in the variability of annual runoff ndashJ Hydrol250 224ndash240

SANDERSON M 1999 The classification of climates fromPythagoras to Koeppen ndash Bull Amer Meteor Soc80669ndash673