Unprecedented...lobal Warming_Chunzai WANG.pdf

Unprecedented Heatwave in Western North America during Late Juneof 2021:Roles of Atmospheric Circulation and Global WarmingChunzai WANG*1,2,3,Jiayu ZHENG*1,2,3,5,Wei LIN2,4,and Yuqing WANG2,41Southern Marine Science and Engineering Guangdong Laboratory(Guangzhou),Guangzhou 511458,China2State Key Laboratory of Tropical Oceanography,South China Sea Institute of Oceanology,Chinese Academy of Sciences,Guangzhou 510301,China3Innovation Academy of South China Sea Ecology and Environmental Engineering,Chinese Academy of Sciences,Guangzhou 510301,China4University of Chinese Academy of Sciences,Beijing 100049,China5State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics,Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing 100029,China(Received 18 March 2022;revised 12 June 2022;accepted 30 June 2022)ABSTRACTAn extraordinary and unprecedented heatwave swept across western North America(i.e.,the Pacific Northwest)in lateJune of 2021,resulting in hundreds of deaths,a massive die-off of sea creatures off the coast,and horrific wildfires.Here,we use observational data to find the atmospheric circulation variabilities of the North Pacific and Arctic-Pacific-Canadapatterns that co-occurred with the development and mature phases of the heatwave,as well as the North America pattern,which coincided with the decaying and eastward movement of the heatwave.Climate models from the Coupled ModelIntercomparison Project(Phase 6)are not designed to simulate a particular heatwave event like this one.Still,models showthat greenhouse gases are the main reason for the long-term increase of average daily maximum temperature in westernNorth America in the past and future.Key words:heatwave,climate change,atmospheric circulation pattern,Pacific NorthwestCitation:Wang,C.Z.,J.Y.Zheng,W.Lin,and Y.Q.Wang,2023:Unprecedented heatwave in western North Americaduring late June of 2021:Roles of atmospheric circulation and global warming.Adv.Atmos.Sci.,40(1),1428,https:/doi.org/10.1007/s00376-022-2078-2.Article Highlights:The North Pacific and Arctic-Pacific-Canada patterns co-occur with the development and mature phases of the heatwavein western North America.The North America pattern coincides with the decaying and eastward movement of the heatwave in western NorthAmerica.Greenhouse gases are a main reason for the average daily maximum temperature increase in the past and future.1.IntroductionWestern North America(i.e.,the Pacific Northwest ofthe United States and Canada),known for its temperateweather in June,normally has average high temperatures inthe comfortable 18C to 24C range.However,westernNorth America experienced an extreme heatwave in lateJune and early July of 2021,resulting in record temperaturesin many places.In particular,Lytton,British Columbiabroke Canadas all-time record on 30 June when the tempera-ture topped 49.5C.This heatwave is also responsible for hun-dreds of deaths,a massive die-off of sea creatures off thecoast,and a spate of horrific wildfires(https:/en.wikipedia.org/wiki/2021_Western_North_America_heat_wave;Over-land,2021).Investigating and understanding the causes ofthis heatwave are both scientifically and socially important.Some studies indicate that increasing trends in the fre-quency,duration,and intensity of heatwaves have acceleratedon global and regional scales under global warming(Perkins et al.,2012;Perkins-Kirkpatrick and Lewis,2020;Wang et al.,2020).There is increasing evidence that anthro-*Corresponding authors:Chunzai WANG,Jiayu ZHENGEmails:,ADVANCES IN ATMOSPHERIC SCIENCES,VOL.40,JANUARY 2023,1428 Original Paper The Author(s)2023.This article is published with open access at .pogenic forcing,such as the forcing of rising greenhousegases,is having an impact on heatwaves (IPCC,2013,2021;Wang et al.,2020;Seong et al.,2021).In addition toexternal forcing,internal variability may also affect the chang-ing nature of heatwaves.Changes in temperature and heat-waves are obviously related to climate modes and atmo-spheric circulations (Lau and Nath,2012;Li and Ruan,2018;Li et al.,2019).Zheng and Wang(2019)identifiedseven atmospheric circulation patterns and combined thesepatterns along with global warming to explain summer sur-face air temperature variations in the Northern Hemisphere.Horton et al.(2015)showed that although a substantial portionof the observed change in extreme temperature occurrencehas resulted from thermodynamic changes,extreme tempera-ture trends have also been altered by recent changes in atmo-spheric circulation patterns.Some studies have also shownthat heatwaves can be modulated by sea surface temperature(SST)anomalies.SST anomalies can remotely influence sum-mer heatwave variabilities over land by generating severe con-vection and triggering atmospheric teleconnections(Wu etal.,2012;Wang et al.,2017;Chen et al.,2019;Deng et al.,2019).This paper uses data from observations and climate mod-els to study physical processes that may be responsible forthe Pacific Northwest heatwave that occurred in late June of2021 from the perspectives of internal variability and externalforcing.Section 2 introduces the datasets,climate models,and methodologies used in this paper.Section 3 shows theresults analyzed from observational data and climate modeloutputs.Finally,section 4 provides a summary and discus-sion.2.Data and methods2.1.Observational dataThe daily atmospheric dataset used in this study is fromthe National Center for Environmental Prediction/NationalCenter for Atmospheric Research(NCEP/NCAR)reanalysisI,which includes surface air temperature at the sigma-995level,geopotential heights,horizontal and vertical winds ona 2.5 2.5 grid(Kalnay et al.,1996).To capture atmo-spheric circulation patterns associated with the Pacific North-west heatwave,the analyses were performed from19792021 since the satellite data were introduced in thelate 1970s(Sturaro,2003).The National Oceanic and Atmo-spheric Administration(NOAA)Optimum Interpolation SeaSurface Temperature (OISST),which is v2.0 from19822015 and v2.1 from 201621,is used to detect the rela-tionship between atmospheric circulation and SST(Reynolds et al.,2007).The Global Historical Climatology Network-Daily(GHCND)dataset is also used in this study(Menne et al.,2012).Daily maximum temperature(Tmax)data at the surfacein a region(4060N,100130W)of western North Amer-ica from 15 June to 15 July during 19512021 are selected.The Tmax anomalies are calculated by using the period of196190 as the climatology.2.2.Climate modelsModel-simulated daily near-surface air maximum temper-ature(“tasmax”)from the Coupled Model IntercomparisonProject Phase 6(CMIP6)is also used in this study.Historicalsimulations are performed to represent the response to allexternal forcing(ALL)by using 17 historical simulationsfrom 7 models(Table 1).The time period used in this studyfor the historical simulations is 19512014.In addition,17simulations from 7 models are used for projections underShared Socioeconomic Pathways scenarios(SSP2-4.5)dur-ing 20152100 (Table 1).The Detection and AttributionModel Intercomparison Project (DAMIP)is a part of theCMIP6.DAMIP contains three tiers of simulations,whichcover 14 types of experiments driven by various forcings,such as well-mixed greenhouse-gas-only(GHG)historicalsimulations,anthropogenic-aerosol-only (AER)historicalsimulations,and natural-only(NAT)historical simulations(Gillett et al.,2016).Here,we use 17 historical simulationsfrom 7 models under GHG forcing,AER forcing,and NATforcing during 19512020 from DAMIP(Table 1).Simi-larly,six simulations from two models (CanESM5 andMIROC6)under GHG forcing,AER forcing,and NAT forc-ing from SSP2-4.5 are used for future projections during20212100.Models are excluded if they do not provide theresults of relevant simulations under GHG,AER,and NATTable 1.List of CMIP6 climate models used in this study.Numbers represent the number of members for the ALL,GHG,AER,andNAT forcings.ALLGHGNATAERHISTSSP245HISTSSP245HISTSSP245HISTSSP245ACCESS-CM233333ACCESS-ESM1-533333CanESM533333333FGOALS-g311111IPSL-CM6A-LR33333MIROC633333333MRI-ESM2-011111Sum(Model)17(7)17(7)17(7)6(2)17(7)6(2)17(7)6(2)JANUARY 2023WANG ET AL.15forcings.Following the spatial resolution of observations,all daily maximum temperature data are interpolated into auniform grid(2.5 2.5)before analysis.PR=PALL/PNATFAR=1PNAT/PALLPALLPNATThe probability density functions(PDFs)of the Tmaxanomalies are calculated to estimate the probability of theobserved 2021 heatwave.The PDFs are estimated by the ker-nel smoothing function(Ma et al.,2017).The probabilityratio(PR)and the fraction of attributable risk(FAR)areused to quantitatively assess the contributions of anthro-pogenic influence to the observed 2021 heatwave and aredefined as follows(Stott et al.,2004;Fischer and Knutti,2015;Ma et al.,2017):;.and represent the probability of exceedingthe extreme event like the observed 2021 heatwave underALL and NAT simulations.The 95%confidence intervalsof PR and FAR are estimated by the bootstrap samplingmethod(Ma et al.,2017).2.3.Empirical orthogonal function analysisDaily 300-hPa geopotential height variations duringJuneJuly for the period of 19792021 are adopted in thisstudy to examine atmospheric circulation patterns associatedwith the heatwave in western North America.First,daily300-hPa geopotential height anomalies are calculated relativeto the climatological mean for 19792021.Second,geopoten-tial height anomalies are weighted by latitude(cosine of lati-tude).Third,daily 300-hPa geopotential height anomalies dur-ing JuneJuly for the period of 19792021,a span of 2623days,are used to perform the empirical orthogonal function(EOF)analysis.The EOF modes identify the atmospheric cir-culation patterns,and their corresponding principal compo-nent(PC)time series are defined as the atmospheric circula-tion pattern indices.2.4.Atmospheric wave activity fluxIn this study,atmospheric wave flux is used to demon-strate atmospheric wave activity fluxes associated with atmo-spheric circulation patterns.First,daily 300-hPa stream func-tion anomalies are calculated relative to the climatologicalmean for 19792021.Second,the stream function anomaliesassociated with atmospheric circulation patterns areobtained as the stream function anomalies regressed ontoatmospheric circulation pattern indices.Third,the wave fluxis computed based on the formula given in Takaya and Naka-mura(Takaya and Nakamura,2001):W=pcos2|U|Ua2cos2|()222|+Va2cos2Ua2cos2+Va2|()222|,(1)p where is the normalized atmospheric pressure(pressure/1000 hPa),a is the earths radius,U and V are climatologicalmean zonal and meridional winds,|U|is the speed of the cli-matological mean winds,denotes the stream functionanomalies,and (,)are latitude and longitude,respec-tively.2.5.Extreme fitting methodA linear fitting method is usually used to reconstructatmospheric variable anomalies(Wang et al.,2013).In thisstudy,this method is used to reconstruct anomalies whenthe value of atmospheric circulation pattern indices is rela-tively large,which is termed the extreme fitting method.First,the slope patterns(SLs)are calculated as the fields ofregression coefficients onto the atmospheric circulation pat-tern indices when the atmospheric circulation patternindices are greater than 1.2 or less than 1.2(is the stan-dard deviation).The intercept patterns(INs),which are thefields of intercepts,are also obtained when the atmosphericcirculation pattern indices are greater than 1.2 or less than1.2.To emphasize extreme atmospheric circulations andtheir effects on the heatwave in western North America,theSLs and INs are set to zero when the atmospheric circulationpattern indices are between 1.2 and 1.2.Second,thedaily 300-hPa geopotential height(GHT)and surface air tem-perature(SAT)anomalies are calculated as follows:GHTA(t)=i=1,nGHTA_fit(Xi(t)=i=1,nSLGHT,XiXi(t)+INGHT,Xi,(2)SATA(t)=i=1,nSATA_fit(Xi(t)=i=1,nSLSAT,XiXi(t)+INSAT,Xi.(3)Here,Xi denotes the North Pacific,Arctic-Pacific-Canada,and North America pattern indices.For example,SLGHT,NPindex1.2 is calculated as the regressed daily300-hPa geopotential height anomalies onto the NorthPacific pattern index when the North Pacific patternindex is above 1.2.Thus,daily 300-hPa geopotentialheight and surface air temperature anomalies can be recon-structed using the atmospheric circulation pattern indicesand their corresponding slope and intercept patterns.16UNPRECEDENTED HEATWAVE IN WESTERN NORTH AMERICAVOLUME 403.Results3.1.Development and decaying of the heatwaveThe devastating heatwave was caused by a heat domeof high pressure(or anticyclone)over western North Amer-ica.The heat dome gets its name because hot air is trappedby a high-pressure system;as the hot air is pushed back tothe ground,it heats up even more(Ramamurthy et al.,2017).As shown in Fig.1,the extreme high-temperature anomaliesover western North America were associated with positive300-hPa geopotential height and negative 500-hPa verticalvelocity anomalies(positive omega).These indicate that theanticyclone and downward motion gave rise to air warmingby adiabatic heating during compression of sinking air(Black et al.,2004).In other words,the heat dome acted asa lid on the atmosphere,which trapped the hot air trying toescape and warmed it even more as it sank.The stronger theheat dome gets,the hotter the near-surface temperature is,and vice versa.In addition to the anticyclone and sinking motion,theheatwave is also associated with the wind change in thetropopause.In a normal condition,a narrow band of verystrong westerly air currents near the altitude of thetropopause,called the jetstream,is relatively flat in the middlelatitudes.During the heatwave,the flat jetstream is distortedinto a wavy pattern and even an omega shape at the peak ofthe heatwave(Fig.1d).The heat dome first became evident on 24 June and grad-ually vanished after 2 July(Fig.1).The positive 300-hPageopotential height anomalies near the coast of North Amer-ica were aligned with the negative geopotential height anoma-lies in the North Pacific and the Arctic,suggesting that theFig.1.The Pacific Northwest heatwave from 24 June to 2 July 2021.(a)Daily surface air temperature anomalies at sigma-995 level(shading;units:C).(b)Daily 300-hPa geopotential height anomalies(shading;units:10 gpm).(c)Daily 500-hPapressure vertical velocity anomalies (shading;positive values denote downward motion;units:102 Pa s1).(d)Daily300-hPa wind speed(shading;units:m s1)and horizontal wind field(vectors;units:m s1).JANUARY 2023WANG ET AL.17heat dome was related to the atmospheric variations in thewest of and north of North America.As shown later by atmo-spheric wave activity flux,the heat dome of western NorthAmerica in late June of 2021 did originate from both theNorth Pacific and the Arctic.To better understand the formation and development ofthe heat dome,we perform an empirical orthogonal function(EOF)analysis of the 300-hPa geopotential hei