Welcome to Scribd. Sign in or start your free trial to enjoy unlimited e-books, audiobooks & documents.Find out more
Download
Standard view
Full view
of .
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
Emerging regional climate change signals for Europe under varying large-scale circulation conditions. Erik Kjellström1, Peter Thejll, Markku Rummukainen, Jens H. Christensen, Fredrik Boberg, Ole B. Christensen, Cathrine Fox Maule. CLIMATE RESEARCH Clim Res, Vol. 56: 103–119, 2013, doi: 10.3354/cr01146; Published online March 12

Emerging regional climate change signals for Europe under varying large-scale circulation conditions. Erik Kjellström1, Peter Thejll, Markku Rummukainen, Jens H. Christensen, Fredrik Boberg, Ole B. Christensen, Cathrine Fox Maule. CLIMATE RESEARCH Clim Res, Vol. 56: 103–119, 2013, doi: 10.3354/cr01146; Published online March 12

Ratings:
(0)
|Views: 6|Likes:
Published by Elisabeta Oprisan
A large ensemble of regional climate model projections was investigated regarding if and when they show an emergence of significant climate change signals in seasonal temperature and precipitation within Europe. The influence of the North Atlantic Oscillation (NAO), as simulated in the projections, was investigated. In most parts of Europe, the projections indicate robust emergence of temperature change in the first 2 decades of the 21st century, typically earlier for summer than for winter. For precipitation changes, signals generally emerge much later than for temperature. For Europe as a whole, the precipitation signals tend to emerge some 40 to 60 yr later than the temperature signals.
A large ensemble of regional climate model projections was investigated regarding if and when they show an emergence of significant climate change signals in seasonal temperature and precipitation within Europe. The influence of the North Atlantic Oscillation (NAO), as simulated in the projections, was investigated. In most parts of Europe, the projections indicate robust emergence of temperature change in the first 2 decades of the 21st century, typically earlier for summer than for winter. For precipitation changes, signals generally emerge much later than for temperature. For Europe as a whole, the precipitation signals tend to emerge some 40 to 60 yr later than the temperature signals.

More info:

Published by: Elisabeta Oprisan on Oct 14, 2013
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

07/27/2015

pdf

text

original

 
CLIMATE RESEARCHClim Res
Vol. 56: 103–119, 2013
doi: 10.3354/cr01146
Published online March 12
1. INTRODUCTION
Climate projections for the 21st century show suc-cessive and substantial changes in regional climateaspects related to global warming (e.g. Christensenet al. 2007). Large-scale patterns such as a largerwarming over the continents than over the oceans,maximum warming in the Arctic region and a gen-eral amplification of the hydrological cycle are com-mon to all projections. However, there are also im-portant differences. On the global scale, differencesbetween projections are related to uncertainties inthe climate forcing itself (emission pathways), the climate sensitivity and, especially for the early part ofthe century, internal variability of the climate system(Hawkins & Sutton 2009). The latter is even more ofan issue on regional scales than on the global scale(Hegerl & Zwiers 2011) and thus is an importantsource of uncertainty in understanding regional-scale climate change signals. These uncertaintieswarrant careful consideration with ensemble ap-proaches that enable characterization of both thecommon features in the projections and the uncer-tainties. This includes an improved understanding ofthe role of the large-scale internal variability as itmay severely mask the externally driven climate
© Inter-Research 2013 · www.int-res.com*Email: erik.kjellstrom@smhi.se
Emerging regional climate change signalsforEurope under varying large-scalecirculation conditions
Erik Kjellström
1,2,
*, Peter Thejll
3
, Markku Rummukainen
1,4
, Jens H. Christensen
3
,Fredrik Boberg
3
, Ole B. Christensen
3
, Cathrine Fox Maule
3
1
Swedish Meteorological and Hydrological Institute, 60176 Norrköping, Sweden
2
Department of Meteorology, Stockholm University, 10691 Stockholm, Sweden
3
Danish Meteorological Institute, 2100 Copenhagen, Denmark
4
Centre for Environmental and Climate Research, Lund University, 22362 Lund, Sweden
ABSTRACT: A large ensemble of regional climate model projections was investigated regardingif and when they show an emergence of significant climate change signals in seasonal tempera-ture and precipitation within Europe. The influence of the North Atlantic Oscillation (NAO), assimulated in the projections, was investigated. In most parts of Europe, the projections indicaterobust emergence of temperature change in the first 2 decades of the 21st century, typically earlierfor summer than for winter. For precipitation changes, signals generally emerge much later thanfor temperature. For Europe as a whole, the precipitation signals tend to emerge some 40 to 60 yrlater than the temperature signals. In some sub-regions, robust signals for precipitation are notfound within the studied period, i.e. until 2100. Some sub-regions, notably the Mediterranean areaand Scandinavia, show different behaviour in some aspects compared to the ensemble-basedresults as a whole. NAO has some influence on the temperature change signals, which emergeearlier in winter for some models and regions if NAO is accounted for. For summer temperatures,the influence of NAO is less evident. Similarly, for precipitation, accounting for NAO leads to anearlier emergence in some regions and models. Here, we find an impact for both summer and winter.KEY WORDS: Climate change · Emerging trends · Europe · Regional climate models · NAO
Resale or republication not permitted without written consent of the publisher 
O
PEN
EN
 
A
CCESS
CESS
 
Clim Res 56: 103–119, 2013
change signal if not properly addressed in a chang-ing climate, where this variability may undergo itsown changes.Existing ensembles of regional climate projectionsfor Europe include those based on global (e.g. Meehlet al. 2007) and regional climate models (e.g. Chris-tensen & Christensen 2007, Haugen & Iversen 2008,Kjellström et al. 2011, Déqué et al. 2012). The resultsindicate robust strong warming trends that are sea-sonally most pronounced in Northern Europe in win-ter and in Southern Europe in summer. Also, thehydrological climate is expected to change with gen-erally more precipitation in the north, especially inwinter, and less precipitation in the south.The projected changes in temperature and precip-itation emerge gradually over time. Detection of con-temporary temperature and precipitation changeshas been achieved on global and continental scales(Hegerl et al. 2007, Stott et al. 2010). The signal-to-noise ratio is more of a constraint on smaller scales.The noise concerns internal variability that mayamplify or suppress the long-term trends and subse-quently complicate early detection of emerging cli-mate change signals (Déqué et al. 2007, Kendon et al.2010, Diffenbaugh et al. 2011, Hawkins & Sutton2012). Giorgi & Bi (2009), Diffenbaugh et al. (2011)and Hawkins & Sutton (2012)showed that in globalclimate model (GCM) projections, regional-scaleprojected changes emerge over different time peri-ods, depending on the region. In many cases, how-ever, definite signals, not least in temperature, areexpected during the next 1 to 2 decades. Forinstance, Sheffield & Wood (2008)showed that airtemperature changes are detectable within the firsthalf of the 21st century at annual and seasonal timescales in the Mediterranean region based on an en-semble of GCMs. A similar result was deduced byChristensen et al. (2007). Precipitation changes,where appropriate, generally become discernible be-yond the emergence of temperature-change signals(e.g. Christensen et al. 2007, Mahlstein et al. 2012),which is in line with the generally lower signal-to-noise ratio in precipitation compared to temperature.In the present study, we look into if and when sta-tistically significant changes in seasonal mean pre-cipitation and temperature could be expected toemerge in Europe. This is interesting for example foradaptation planning and policy (e.g. Joshi et al.2011). We also consider sub-regions of Europe, andour point of departure is an ensemble of regional cli-mate model (RCM) projections from the ENSEM-BLES project (van der Linden & Mitchell 2009),which differs from the above-mentioned studiesbased on GCMs. RCMs offer a better resolved simu-lation of temperature and precipitation and how theirdistributions, and possibly changes, are affected bytopography (Rummukainen 2010, Feser et al. 2011).We also investigate how the emergence of regionalclimate change signals is influenced by variations inthe North Atlantic Oscillation (NAO). The latter ismotivated by the fact that large-scale circulationmodes in the North Atlantic/Eurasian region have aprofound impact on the inter-annual and decadalvariability in the region, and of these, the NAO is aprime example with its positive (negative) phaseleading to unseasonally mild (cold) winters in North-ern Europe (Hurrell 1995). Possible future change inNAO is an important driver for changes in regionalclimate properties in central Europe (van Olden-borgh et al. 2009), and it has a profound influence onthe regional inter-annual variability.
104Number Institute RCM GCM 1
a
C4I RCA3
b
HadCM3Q16
e
2
a
CNRM RM5.1 ARPÈGE3 DMI HIRHAM5
c
ARPÈGE4 DMI HIRHAM5
c
ECHAM55 DMI HIRHAM5
c
BCM6 ETHZ CLM
d
HadCM3Q0
e
a
GKSS CLM
d
IPSL
8 ICTP REGCM3 ECHAM59
a
KNMI RACMO2 ECHAM510
a
METO-HC HadRM3Q0
e
HadCM3Q0
e
11 METO-HC HadRM3Q3
e
HadCM3Q3
e
12 METO-HC HadRM3Q16
e
HadCM3Q16
e
13 METNO HIRHA
c
BCM 14 METNO HIRHA
c
HadCM3Q0 
e
15 MPI-M REMO ECHAM5
16 
a
OURANOS MRCC4.2.1 CGCM
17
a
SMHI RCA3
b
BCM18
a
SMHI RCA3
b
HadCM3Q3
e
19 SMHI RCA3
b
ECHAM5
20 UCLM PROMES HadCM3Q0 
e
21 VMGO RRCM HadCM3Q0 
ea
Used for deriving the NAOI associated with the GCMs
b
The RCA3 at C4I and at SMHI was operated in slightlydifferent configurations
c
The HIRHAM version was different at DMI and METNO
d
The CLM version used at ETHZ and GKSS was different.Furthermore, the latter’s simulation started in 1963,rather than 1961 or earlier
e
Three different versions of the Hadley Centre modelswere used: Q0 (i.e. the reference version), Q3 with lowclimate sensitivity and Q16 with high climate sensitivityTable 1. Regional climate model (RCM) simulations in thepresent study. Models denoted in
italics 
stop at 2050 and arenot included in the analysis of year of change unless noted inthe text
 
Kjellström et al.: Emerging regional climate change in Europe
2. DATA AND METHODS2.1. Regional climate model data
We used monthly mean temperature, precipitationand mean sea level pressure (MSLP) fields from21RCM simulations from the ENSEMBLES data-base at DMI (ENSEMBLES 2012) (Table 1). Webased most of our analysis on the 15 simulationsthat cover the full 1961−2099 period and use theremaining 6 ending in 2050 to discuss the uncer-tainty in the results. The ENSEMBLES simulationswere made with 11 different RCMs. Some of thesewere run with different set-ups. For instance, theHadley Centre RCM, HadRM3, was used in 3 dif-ferent configurations with different parameter set-tings from a perturbed physics ensemble (Collins etal. 2011). The RCMs’ boundary conditions werederived from 8 different GCMs. The anthropogenicclimate forcing followed the A1B emission scenario(Naki
´
cenovi
´
c & Swart 2000). In some RCMs, directand indirect effects on the climate from sulphateaerosols were included in addition to the effect ofincreased amounts of greenhouse gases (GHG),while in others, only the GHG increases wereaccounted for.
2.2. Observational data
In addition to the RCM data, weused gridded observational tempera-ture and precipitation data from E-OBS5.0 (Haylock et al. 2008) for com-paring spatial patterns and temporaltrends. The E-OBS data are availableat the same rotated 25 × 25 km gridthat is used by most of the ENSEM-BLES RCMs. NAO was characterizedwith a station-based seasonal meanNAO index (NAOI) (NAO 2012).
2.3. Methods
Temperatureandprecipitationdatawereaveragedforthe9regionsshownin Fig. 1for both the RCMsand the E-OBS data. Anomalies werethen calculated by subtracting theaverage in the period 1961−1990(except for Model 7, for which 1963−1990 was used as no data wereavailable prior to 1963). We calculated when thetemperature and precipitation deviated from the1961−1990 reference period for winter (Decemberto February, DJF) and summer (June to August,JJA) in each region. These deviations were ex-pressed as a probability given the standard devia-tions of 2 samples; one from the reference periodcovering 30 yr and the other from a sliding windowof length of 30 yr. The probability that these 2 sam-ples were drawn from the same population wasthen calculated using Student’s means test. If thehypothesis that they were drawn from the samepopulation could be rejected at the 1% level (a 1%chance that the rejection was due to a chanceoccurrence), then the centre year of the sampleinterval for that model (cf. Fig. 2) was assigned asthe time when the climate change signal emerged.Temporal autocorrelation was not accounted for inthe calculation of the
-test statistics. This meansthat the estimated climate change signal mayemerge too early. However, the autocorrelation forseasonal mean temperature and precipitation in theRCM data is relatively small (data not shown),implying that the inferred error is small. Also, wenote that the autocorrelation is of similar size fortemperature and precipitation, implying that thetime of emergence (ToE) can be compared between
105
Fig. 1. Areas used in the analysis. The overall European region (EU) encompassesthe entire domain shown in the figure. The colours represent the orography asprescribed in the HIRHAM5 model at DMI. See Table 2 for abbreviations

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->