AMO-Forced Regional Processes Affecting Summertime Precipitation Variations in the Central

16 Pages·2017·4.57 MB·English

University of Nebraska - Lincoln
DigitalCommons@University of Nebraska - Lincoln
Papers in the Earth and Atmospheric Sciences Earth and Atmospheric Sciences, Department of
2013
AMO-Forced Regional Processes Affecting
Summertime Precipitation Variations in the
Central United States
Michael C. Veres
University of Nebraska-Lincoln, [email protected]
Qi Hu
University of Nebraska-Lincoln, [email protected]
Follow this and additional works at:http://digitalcommons.unl.edu/geosciencefacpub
Veres, Michael C. and Hu, Qi, "AMO-Forced Regional Processes Affecting Summertime Precipitation Variations in the Central United
States" (2013).Papers in the Earth and Atmospheric Sciences. 397.
http://digitalcommons.unl.edu/geosciencefacpub/397
This Article is brought to you for free and open access by the Earth and Atmospheric Sciences, Department of at DigitalCommons@University of
Nebraska - Lincoln. It has been accepted for inclusion in Papers in the Earth and Atmospheric Sciences by an authorized administrator of
DigitalCommons@University of Nebraska - Lincoln.
276 JOURNAL OF CLIMATE VOLUME26
AMO-Forced Regional Processes Affecting Summertime Precipitation
Variations in the Central United States
MICHAELC.VERES
DepartmentofEarthandAtmosphericSciences,UniversityofNebraskaatLincoln,Lincoln,Nebraska
QIHU
DepartmentofEarthandAtmosphericSciences,andSchoolofNaturalResources,UniversityofNebraska
atLincoln,Lincoln,Nebraska
(Manuscriptreceived16November2011,infinalform30May2012)
ABSTRACT
NumerouspreviousstudieshaveprovidedinsightintotheinfluenceoftheAtlanticmultidecadaloscillation
(AMO) on North American precipitation. However, these studies focused on large-scale processes, and
additionalstudiesareneededtogainunderstandingoflocalandregionalprocessesthatdevelopindifferent
phasesoftheAMOandsubstantiateitsinfluencesonprecipitation.Inthisstudy,theWeatherResearchand
Forecasting(WRF)regionalmodelisusedtoexamineAMO-forcedlocalandregionalprocessesandhow
theyhaveaffectedsummertimeprecipitationvariationinthecentralUnitedStates.
While moisture transport and convergence by the Great Plains low-level jet have been recognized as
necessary conditions for summer precipitation, model simulations show similar low-level moisture flux
convergenceinthecentralUnitedStatesbetweenthecoldandwarmphasesoftheAMO.However,therewas
astrongmoisteninginthelowertroposphereduringtheAMOcoldphase,makingtheatmospheremore
unstableforconvectionandprecipitation.Thesourceofthemoisturewasfoundtobeastrongpositivesurface
evaporation–precipitation feedback initiated and sustained by increased relative vorticity along a frontal
zone.Alongthefrontalzone,isentropicstretchingoftheupper-levelatmosphereandcycloniccirculation
anomaliesincreasedtherelativevorticityduringtheAMOcoldphase,providingthedynamicsupportneeded
toreleasethelow-levelmoistinstabilityandproducetheincreasedprecipitation.Theseresultsindicatethat
thedynamicsofthecirculationintheAMOcoldphaseplayedkeyrolestoorganizeregionalvorticitypro-
cessesthatfurthersustainedacouplingofprecipitationandthesurfaceevaporationandperpetuatedthe
precipitation.
1. Introduction continental-scale circulations, such as those described
by general circulation models (GCMs). It is these re-
In the central United States, summer precipitation
gional processes that influence regional weather and
anomalies strongly affect agricultural production, the
climate.
environment,andsocietybydamagingfloodsorstraining
Several regional processes have been recognized as
watersuppliesandenhancingtheriskofwildfiresduring
contributing to precipitation anomalies in the central
drought. It would be possible to mitigate some of the
UnitedStates.Thelow-levelsoutherlymoistflowfrom
consequencesof the precipitation anomalies ifthey can
theGulfofMexicoisamajorone(e.g.,Rasmusson1967;
be better understood and predicted accurately. To do
Arrittetal.1997;Moetal.1997,2009).Asthesoutherly
this would require a more extensive understanding
moistflowisconcentratedinachannelfromtheGulfof
of the physical processes causing the precipitation
Mexico, it is frequently referred to as the Great Plains
anomalies. To gain such an understanding, regional-
low-leveljet(LLJ;e.g.,Bonner1968).Ithasbeenshown
scalecirculationsmustbeexaminedinadditiontothe
thatvariationsofthegeographicallocationandintensity
oftheLLJaredirectlyrelatedtothevariationsinspring
and summer precipitation in the central United States
Corresponding author address: Dr. Qi Hu, 707 Hardin Hall,
(e.g.,BellandJanowiak1995;Arrittetal.1997;Higgins
UniversityofNebraskaatLincoln,Lincoln,NE68583-0987.
E-mail:[email protected] etal.1997).TuttleandDavis(2006)furthershowedthat
DOI:10.1175/JCLI-D-11-00670.1
(cid:1)2013AmericanMeteorologicalSociety
1JANUARY2013 VERES AND HU 277
the exit region of the LLJ is a favorable area for sum- AMO is a 60–80-yr variation in the North Atlantic
mertimeconvectiveprecipitationdevelopment. SST (Enfield et al. 2001). Although only about two
Regional processes associated with changes in the complete cycles of the AMO have been observed in
intensityandpositionoftheupper-levelwesterlyjetand modern instrumentation data, the SST variations at
relatedrelativevorticityareothermajorprocesses.For the multidecadal time scale in the North Atlantic
example, during the 1993 floods in the central United have been simulated in long-term GCM simulations
StatesandtheMidwest,thejetstreamwasfoundtobe (e.g., Knight et al. 2005) and observed in climate re-
stronger than average and shifted southward from its constructions using tree-ring data and other proxy
climatological position, and there was strong regional records of the past 7000 yr (Gray et al. 2004; Feng
advection of positive relative vorticity from the north- etal.2011).
westerntothecentralUnitedStates(BellandJanowiak TheimpactofNorthAtlanticSSTforcing onNorth
1995; Trenberth and Guillemot 1996; Mo et al. 1997). American precipitation has been analyzed by using
The southward shift of the stronger jet displaced the observedanomalies(e.g.,Enfieldetal.2001;Schubert
seasonal storm track to the south, making it easier for et al. 2004; Hu and Feng 2008; Feng et al. 2011) and
storms along the storm track to access the low-level GCMsimulationsusingidealizedSSTanomalies(e.g.,
moistureinthesoutherlyflowfromtheGulfofMexico. Sutton and Hodson 2005, 2007; Hu and Feng 2007;
Withthevorticityanomaliesenhancedregionalstorms Fengetal.2008;Huetal.2011).Thegeneralconsensus
contributed to increasing summertime precipitation in isthatwarmerSSTintheNorthAtlanticOceanduring
the central United States. Trenberth and Guillemot the warm phase of the AMO produces below average
(1996) also identified an inverse process during the se- summertime precipitation in the central United States.
veredroughtof1988. Schubert et al. (2004) attribute the 1930s ‘‘Dust Bowl’’
While revealing the contributing factors influenc- era drought to warmer North Atlantic SST. In general,
ing precipitation anomalies andextreme precipitation the warmer SST weakens the North Atlantic sub-
events in the central and the U.S. Midwest and their tropical high pressure (NASH) and modifies the LLJ
interrelationshipwithanomaliesinotherregions,these onthewesternflankoftheNASHandcirculationover
studies emphasized that local and regional processes North America (Dong et al. 2006; Hu et al. 2011).
are essential for us to gain detailed understanding of During the cold phase of the AMO, summer pre-
regional precipitation variations. These processes con- cipitation in the central United States is generally
stitute the large-scale circulation anomalies driven by foundtobeaboveaverage(SuttonandHodson2005;
internal and external forcings. Understanding these HuandFeng2008).
processes and their relationship with the forcings can Interactions of the AMO with other SST forcings,
therefore improve our ability to describe these pro- suchasElNin˜oandLaNin˜a,andtheireffectsonNorth
cesses and to predict regional precipitation and ex- American precipitation also have been examined re-
tremeevents. cently(Schubertetal.2009;Moetal.2009;HuandFeng
SomeoftheforcingsofthecentralU.S.summertime 2012).Allthesestudiesshowverydryconditionsinthe
precipitation variations have been attributed to anom- central and south-central United States in La Nin˜a
alies of the sea surface temperatures (SSTs) of the duringthewarmphaseoftheAMO,aresultconsistent
Pacific and Atlantic Oceans (e.g., Namias 1959; 1965; with previous studies showing enhanced droughts in
1969; Namias et al. 1988; Bjerknes 1964; Frankignoul those regions during La Nin˜a (Hoerling and Kumar
1985;LauandNath1994,1996;LatifandBarnett1994; 2003)orinthewarmphasesoftheAMO(HuandFeng
Seageret al.2000; Kushnir et al. 2002;Ting and Wang 2008). There are differences, however, for precipita-
1997; Hu and Feng 2001, 2008; Mo et al. 2009; Wang tion anomalies in the other joint phases of the AMO
etal.2008,2010;Huetal.2011;Fengetal.2011).Huand and ENSO. For example, the results in Mo et al.
Feng (2001) showed that the North Pacific SST anom- (2009)suggestthatabove-averageprecipitationinNorth
alies are regulating the impacts of tropical SST varia- America develops in El Nin˜o years in the cold phase
tionsassociatedwithElNin˜o–LaNin˜aonsummertime of the AMO, while the results in Hu and Feng (2012)
precipitation variations in the central United States. show a near-average summertime precipitation in
McCabeetal.(2004)furthersuggestedthatthePacific North America in such a situation. As suggested in
decadal oscillation (PDO) accounts for 24% of the Hu and Feng (2012), these differences could have re-
droughtvarianceinthecontinentalUnitedStates.They sulted from the differences in analyses (e.g., effects
also showed that the Atlantic multidecadal oscillation on summertime precipitation versus annual precipi-
(AMO; Mestas-Nunez and Enfield 1999; Kerr 2000) tation) between those studies. Those differences point
accounts for 28% of the U.S. drought variance. The toneedsforfurtherinvestigationstoclarifythemand
278 JOURNAL OF CLIMATE VOLUME26
for better understanding of the effects of ENSO and gridded with resolutions of 2.58 3 2.58 in latitude and
theAMO. longitudeandcontain17pressurelevelsinthevertical
While these studies are advancing our knowledge of direction. Even though the NCEP–NCAR reanalysis
individualandcollectiveeffectsoftheAMOandENSO data are not observations per se, they are derived from
onatmosphericcirculationandprecipitation,theyhave and are tightly constrained by observations. For this
broughtusonlythelarge-scaleeffectsofthoseforcings. study,themeanmonthlywindsandgeopotentialheights
Thelocalandregionalprocessesthatdevelopwithinthe fromthesedatawillbeanalyzedatthreepressurelevels:
large-scale circulation environment driven by those 850,500,and300 hPa.
forcings are essential for causing the observed pre- Another dataset used in this research is from the
cipitation anomalies and need to be understood for outputsoftheNCARCommunityAtmosphereModel,
improvingpredictionsofregionalprecipitationandex- version 3.1 (CAM3.1) experiments performed by Hu
treme events. This current research is focusing on the etal.(2011).Intheirexperiments,theCAM3.1wasrun
regional processes in North America driven by the withconstrainedglobalSSTforwarmandcoldphasesof
AMOandintendstoprovidein-depthunderstandingof theAMO.TheresultspresentedinHuetal.(2011)were
the AMO forcing on the summertime precipitation based on the first 20 yr of model integrations for each
variationsinthecentralUnitedStates.Themethodused phaseoftheAMO,butthesemodelrunswereextended
in this study is a regional model, which allows us to to 50 yr in length. It is this extended dataset that was
simulateregionalandmesoscaleprocessesthataretoo used in part for our study. The resolution of the data
smallinscaletobesufficientlyresolvedanddescribedin producedbyHuetal.(2011)isT42,equivalentto2.883
coarser-resolutionGCMs.Itispotentiallythroughthese 2.88 in latitude and longitude resolution, and data are
regional and mesoscale processes that the large-scale availableat6-hourlyintervals.UnliketheNCEP–NCAR
forcings,suchastheAMO,arecapitalizedinregionsto reanalysis,whichincorporatesalltheforcingsthatoccur
produce the observed precipitation anomalies and ex- intherealatmosphere,theAMO-only-forcedGCMdata
tremes. provide a comprehensive dataset of atmospheric vari-
The regional model and data used in this study are ables that vary only by processes driven by the North
described in detail in the next section (section 2). In AtlanticSSTanomaliesrelatedtotheAMO.
section 3, results from testing and validation of the
b. Methods:Theregionalmodel
model are presented and discussed. These results pro-
vide support for applying this model to study the re- ToidentifytheAMO-forcedregionalandmesoscale
gionaleffectsoftheAMO.Afterthevalidation,model processes and mechanisms over the central United
experimentsandresultsaredescribedandevaluatedin States that produce the observed and modeled boreal
section 4 and, based on these results, a mechanism is summer [June–August (JJA)] precipitation anomalies
proposedtoexplaindevelopmentofthemajorregional we used the regional model, the Advanced Research
processes by which the AMO influences central U.S. WeatherResearchandForecastingmodel(ARW-WRF),
summertime precipitation. Major conclusions are con- version3.1(Skamarocketal.2007).
tainedinsection5. Twonesteddomainswereusedinthemodelsetting,an
outerdomainandaninnerdomain(Fig.1).Thefeedback
option between the inner and outer domains was not
2. Dataandmethods used.Intheearlystagesofthisstudy,thefeedbackoption
wasused,andtheresultshowedprecipitationofmagni-
a. Data
tudes much weaker than the observed. After further
The observed monthly precipitation data developed evaluations of the feedback process, it was deemed in-
attheGlobalPrecipitationClimatologyCentre(GPCC; capabletodescribetheinteractions of the domainsand
Beck et al. 2005; Rudolf and Schneider 2005; Rudolf thereforewasremoved.
et al. 2003, 2005) were used in this study. The GPCC The outer domain of the model has a resolution of
dataareagloballygriddeddatasetat0.5830.58latitude 48km348kmandcontainsnearlyallofNorthAmerica,
and longitude resolution and span from 1951 to 2004. from southern Mexico (around 158–208N) to northern
Also, monthly and 6-hourly atmospheric data from Alaska and the Queen Elizabeth Islands (around 658–
the National Centers for Environmental Prediction– 758N).Thewesternandeasternboundariesoftheouter
National Center for Atmospheric Research (NCEP– domainvarywidelyasthenativemapprojectionforthe
NCAR) reanalysis (Kistler et al. 2001; Kalnay et al. domain is Lambert conformal. In general, the western
1996) were used. The NCEP–NCAR reanalysis data and eastern boundaries lay approximately 208 of longi-
are from 1948 to the present. The data are globally tudeoffthecoastsofthecontinentalUnitedStates.The
1JANUARY2013 VERES AND HU 279
surface physics options and the Kain–Fritsch (KF) para-
meterization scheme for atmospheric convection. Model
outputisat3-hintervals.
3. Modelvalidation
Before using WRF to examine the AMO-forced re-
gional processes and precipitation, we need to verify
that the model is capable of producing realistic atmo-
sphericcirculationsandprecipitationinthestudyregion
duringtheAMO.
TheNCEP–NCARreanalysisdatawereusedtoforce
theregionalmodelforfivespecifiedsummersfromeach
phaseofthecurrentAMOcycle,foratotalof10sum-
mer seasons. According to Hu and Feng (2008) (see
theirFig.6),thecoldphaseoftheAMOwasfrom1961
to1990andthewarmphasewasfrom1991to2009.In
FIG.1.Thetwodomainsusedinthisstudy.Theouterdomain selectingthe10 yrforthevalidation,weusedtheresults
includestheentireimageandtheinnerdomain(d02)isboundedby of Hu and Feng (2008), which showed that the central
theheavyblackbox.Thethreelinesinsidetheinnerdomainarethe United States (Kansas, Nebraska, Iowa, and Missouri)
threecrosssectionsdiscussedinsection4.
experiences increased (decreased) precipitation during
thecold(warm)phaseoftheAMO.However,asRuiz-
primarypurposeforusinganouterdomainmuchlarger BarradasandNigam(2005)showedthecorrelationbe-
than the inner domain (Fig. 1) is to provide adequate tween the observed and the NCEP–NCAR reanalysis
distancebetweentheboundariesofthetwodomainsso precipitation is weak in the Great Plains and, when
toprevent,asmuchaspossible,theboundaryforcingon compared to the observations, the NCEP–NCAR re-
theouterdomainfromhavinglargeandsomewhatdirect analysisdatashowabiastowardincreasedsummerpre-
effectontheboundariesoftheinnerdomain.Bylimiting cipitation during the previous 30yr. Recognizing this
thisboundaryeffect,themodelwillhaveagreaterfree- bias,wefocusedonyearswhenboththeobservationsand
domindevelopingtheinternalphysicalprocessesinre- the NCEP–NCAR reanalysis were indicating increased
sponsetotheAMOforcing. (decreased) precipitation during the cold (warm) phase
The inner domain has a finer resolution of 12 km 3 oftheAMO. Oftheseyears,theoneswiththegreatest
12 km, which allows for development of small-scale precipitation anomalies were selected. The low correla-
surface and atmospheric processes within the domain. tion between the observations and the NCEP–NCAR
The inner domain encompasses nearly all of the conti- reanalysisalsorequiredthatthepresenceofElNin˜oor
nentalUnitedStateseastoftheRockyMountains,with La Nin˜a not be a deciding factor. The years selected
onlytheomissionofpartofeasternNewEngland.The throughthisprocessforthemodelvalidationwere1965,
southern boundary starts near 258N and includes the 1966,1979,1981,and1986fortheAMOcoldphaseand
northern half of the Gulf of Mexico. The northern 1999–2003fortheAMOwarmphase.
boundaryisapproximatelyparalleltotheU.S.–Canadian Foreachselectedyear,WRFwasrunfrom16Mayto
borderinthewesternhalfofthedomainandincludes 1 September in each simulation. The first 2 weeks of
the far southern reaches of the Ontario and Quebec each model run was discarded as model spinup. These
provincesofCanadaintheeasternhalfofthedomain. resultswerethencomparedtoobservationstoevaluate
The western and eastern boundaries of the inner do- the effectiveness of the model at simulating the ob-
mainareapproximatelyat1108Winthewestandalong servedconditions.
the Atlantic coast of the continental United States in Wenotethatbecausefiveyearsinthewarmphaseand
theeast(708–788W).Expanding theboundariesofthe five years in the cold phase of the current AMO cycle
inner domain beyond these limits was not feasible, as were used in the validation simulations, there was no
the boundaries were near the limits of the available ‘‘control run.’’ As such, the anomalies were calculated
computationcapabilities. byremovingtheaverageofall10simulationsfromeach
All WRF runs performed in this study used identical simulated year. The equal number of years from the
domainsandmodelparameters.Themodelwasrunusing cold and warm phases resulted in mirror images for
the Noah land surface model and Monin–Obukhov anomalies for the cold and warm phases of the AMO.
280 JOURNAL OF CLIMATE VOLUME26
FIG.2.Mean(10yr)JJAprecipitationfor(a)NCEP–NCARreanalysis–forcedWRFsimulationand(b)observations.(c)Simulatedand
(d)observedprecipitationanomaliesinthecoldphaseoftheAMO.Unitsareinmillimetersperday.
The following discussions primarily focus on the anom- Themodel-simulatedandobservedprecipitationsare
aliesinthecoldphaseoftheAMO. shown in Fig. 2. Comparisons of these results show
Also in the following discussions, we use ‘‘central simulated precipitation consistent with observations in
United States’’ for the region encompassing Kansas, mostofthestudyregioninboththe10-yrmeans(Figs.
Nebraska, Missouri, and Iowa; ‘‘south-central United 2aversusFig.2b)andcoldphaseoftheAMO(Fig.2c
States’’ for Texas, Oklahoma, western Louisiana, and versus Fig. 2d). The precipitation magnitudes are also
Arkansas;‘‘north-centralUnitedStates’’forNorthand consistentthroughoutmuchofthemodeldomain(Fig.
South Dakota and Minnesota; and ‘‘the Midwest’’ for 2a versus Fig. 2b). During the AMO cold phase, how-
Wisconsin, Illinois, Indiana, Michigan, and Ohio. In ever,thereisadisagreementinprecipitationanomalies
addition,specificstateswillbementionedwhenneeded in the east and southeast United States (Fig. 2c versus
forclarity. Fig.2d).Whiletheobservationsshowabandofnegative
1JANUARY2013 VERES AND HU 281
anomalies of precipitation from east of Pennsylvania States.Besidesthisminordifference,themodelsimu-
down to Louisiana, the simulations show positive lation is consistent with the observation in the lower
anomalies across that region. In the simulation result, troposphere.
a band of negative precipitation anomalies is found At 500 and 300 hPa, the cold phase geopotential
along the southeast coasts of the United States and height anomalies display an anomalous wave pattern
across Florida. This misplaced dry band also causes in both the observations (Figs. 3b,c) and simulations
discrepancywiththeobservednarrowbandofpositive (Figs.3e,f).Thispatternissimilartotheoneat850 hPa,
precipitation anomalies along the coastal areas and withthemostpositive(negative)anomaliesoccurringin
northern Florida. The displacement of the simulated thesoutheast(northcentral)UnitedStates.
anomaliestowardthecoastalareaislikelytheresultof To summarize, the regional model simulations gen-
the model producing more zonal flow in the southern erally produced similar results in both magnitude and
United States. This zonal flow traps more moisture distribution to observations. The model simulated the
fromtheGulfofMexicointhesouthernUnitedStates, mid-andupper-levelanomalouscirculationsreasonably
likely increasing the precipitation and producing the wellinthestudydomain.Atthelowlevel,theaccuracy
simulatedpositiveanomaliesinMississippi,Louisiana, wasweaker.However,theoverallpattern ofananom-
and Georgiaduring thecoldphase of the AMO. While alous high pressure in the southeastern United States
the displaced anomalies indicate model limitations in andlowpressureoverthecentralUnitedStatesduring
simulating precipitation in the east coastal regions, the thecoldphaseoftheAMOwasreproduced.Themodel
model appears to be reasonably consistent in simulat- alsosimulatedtheobservedincreaseinprecipitationin
ingprecipitationinthecentral,northandnortheastern the central United States during the cold phase of the
United States, as the simulated anomalies in those re- AMO, albeit the agreement between the simulations
gionsareconsistentwiththeobservations. andtheobservationswasweakerintheMidwest.These
Itshouldalsobenotedthatthemodeldoesnotcap- overallperformancesofthemodelshowthatitwasable
ture the observed nocturnal peak in diurnal precipi- todescribemajorphysicalprocessesthatproducedthe
tationintheGreatPlains(Wallace1975).Thisislikely studyphenomena.Additionaldetailsofthedifferences
due to the poor representation of the diurnal cycle between the observation and model simulations and
by the Kain–Fritsch cumulus parameterization scheme more discussions of model ability in describing the
usedintheWRFmodel(Liangetal.2004).However,our phenomenaofinteresttothisstudyweredocumented
analysisofthediurnalcycleofthemodel-simulatedpre- inVeres(2011).
cipitation indicates approximately an equal amount of
precipitation between local daytime (1200–0000 UTC)
4. Modelexperimentsandresults
andnighttime(0000–1200UTC)contributingtothedaily
precipitation.Thesameresultappearedinboththeyears We next used the model in a series of experimental
selectedfortheAMOwarmandcoldphases.Thesere- simulations. These simulations were forced by GCM
sults suggest that the simulated daytime and nocturnal outputs with specified (constant) SST anomalies de-
response to the AMO forcing issimilar and that a shift scribing the AMO in the North Atlantic Ocean and
towardmorenocturnalprecipitationwouldnotalterthe climatologicalSSTelsewhere(Huetal.2011).Thede-
simulateddailytoseasonalprecipitationshowninFig.2c. tailson the SST distributions used to drive the GCM
Inotherwords,theinaccuracyindescribing thediurnal simulations are shown in Fig. 1 of Hu et al. (2011).
cycle of precipitation caused by the KF scheme is not Outputs of five GCM simulation years for the warm
importantinthisstudy. orcoldphaseoftheAMOfromHuetal.(2011)were
Toevaluatethemodelsimulationoftheatmospheric usedtodrivetheWRFexperiments.Thefiveyearsfor
circulations, we compared the simulated and observed the cold (warm) phase of the AMO were selected
windsandgeopotentialheightsat850,500,and300 hPa in a manner consistent with the NCEP–NCAR re-
(Fig. 3). Overall, the model is able to reasonably re- analysisforcing;thatis,thefivewettest(driest)years
producetheobservations.DuringtheAMOcoldphase, in the cold (warm) phase of the AMO were used.
theprimaryfeatureat850 hPaintheobservationsand Because only AMO-driven large-scale circulations
simulations is an anomalous wave pattern in the geo- were used to force the regional model, by analyzing
potential heights (Figs. 3a,d) with the least negative the regional model simulations forced in different
(more negative) anomalies in the southeastern (north phasesoftheAMOwecanidentifytheAMO-driven
central) United States, although the wave is more regional-scale dynamic processes and associated sum-
meridional in the simulations, producing wind anom- mertime precipitation variations in the central United
alies that are more zonal in the south-central United States.
282 JOURNAL OF CLIMATE VOLUME26
FIG. 3. (a)–(c) Observed and (d)–(f) simulated JJA (a),(d) 850-, (b),(e) 500- and (c),(f) 300-hPa wind and geopotential height
anomaliesforthecoldphaseoftheAMOusingthefiveselectedyearsforthecoldphase(1965,1966,1979,1981,and1986).Unitsarein
meterspersecondforwindsandmetersforgeopotentialheight.Referencevectorsare(a),(d)0.5and(b),(c),(e),(f)1ms21.Contour
intervals are (a),(d) 0.5 and (b),(c),(e),(f) 1m with labels every other contour. Solid (dashed) lines indicate positive (negative)
anomalies.
Figures4a,bshowthemodel-simulatedAMO-forced experiences above-average precipitation while the
JJA 10-yr mean and cold phase anomalous precipi- southeastern United States (e.g., Florida) has below-
tation, respectively. Comparisons of Fig. 4a with the average precipitation. Changes in summertime precip-
NCEP–NCARreanalysis–forcedprecipitationinFig.2b itation in these regions between the cold and warm
indicate that the model produced less precipitation in phasesoftheAMO arestatisticallysignificantatthe
thecentralUnitedStates,theMidwest,andparticularly 95% confidence level (shown in the hatched area in
inthesouth-centralUnitedStates.Despitethereduced Fig. 4b). These anomalies display a strong similarity to
magnitudes, the spatial pattern of the mean precipita- the NCEP–NCAR reanalysis–forced model simulations
tioninthestudyregionissimilartotheobserved1971– (Fig.2c),despitethefactthatNCEP–NCARreanalysis–
2000 climatology(Fig. 4c),withprecipitation generally forced simulations have all possible forcings, besides
increasing toward the north and the east. Both the the AMO. This similarity thus suggests that the AMO
simulations and observations display similar regions of influenceisplayingadominantroleinsummertimepre-
higherprecipitationthatextendmeridionallyfromMin- cipitation variations in North America at the multi-
nesotaandWisconsintoaroundtheMissouri–Arkansas decadal timescale.Theoppositeprecipitationanomalies
border and taper off toward the west. The westward weresimulatedbythemodeldrivenbythewarmphase
gradient is stronger in the simulations, but the overall AMOforcing.
patternoutsidethesouth-centralUnitedStatesissimilar. Toexaminetheregionalandlocalprocessesthatmay
Despite the smaller magnitude in the JJA mean pre- have contributed to these contrasting precipitation
cipitation,thesimulatedcoldphaseprecipitationanom- anomalies between the cold and warm phase of the
alies(Fig.4b)agreefairlywellwithpreviouslydescribed AMO, we start with the analysis of the moisture pro-
AMO-forcedpatterns(e.g.,Enfieldetal.2001;Huetal. cesses.Asshowninmanypriorstudies(e.g.,Arrittetal.
2011). Much of the central and eastern United States 1997;Higginsetal.1997),changesinmoisturecontent
1JANUARY2013 VERES AND HU 283
ofthelowertroposphere canhaveastrongimpacton
precipitation. Figure 5 shows theJJA surface–700-hPa
integrated moisture flux divergence in the study do-
main. Interestingly, low-level moisture convergence is
shown in the central United States and neighboring
areasintheGreatPlainsduringboththecold(Fig.5a)
and warm phase (Fig. 5b) of the AMO. The similar
moisture convergence in the central United States in
Figs. 5a,b indicates thegeneral availability of moisture
inthelowertroposphereduringJJAineitherthecoldor
the warm phase of the AMO. The similarities in low-
levelmoistureconvergencebetweenthecoldandwarm
phasesoftheAMOindicatethatdifferentdynamicpro-
cessmustbeactive,suchthatthoseprocessesinthecold
phase of the AMO would make the atmosphere more
efficient in converting moisture into precipitation and
sustaintheincreaseinJJAprecipitation.
Toidentifythoseprocessesthatmaketheatmosphere
more conducive to storm development in JJA, we first
examine the thermodynamic profile of the atmosphere
inthecoldandwarmphasesoftheAMO.Anunstable
profileisessentialforconvectivestorms,whicharethe
primaryformofJJAprecipitationinthecentralUnited
States. Figure 6 shows three vertical cross sections of
moisture anomalies. All three cross sections share
a common southern endpoint in northeast Texas and
diverge toward the north, terminating in Michigan,
Wisconsin,andNorthDakota.Theywillbereferredto
as eastern, central, and western cross sections and are
identifiedinFig.1bysolid,dashed,anddottedstraight
lines, respectively. Intriguingly, all three cross sections
showmoisteninginthelayersbelow700or750 hPaand
dryingaloftduringthecoldphaseoftheAMO.Areversed
moistureanomalyprofileoccursduringthewarmphaseof
theAMO.Themoisteninginthelowertroposphereand
dryingaloftinthecoldphasereducethemoiststaticsta-
bility of the atmosphere, making it more convectively
unstableandpronetoconvectionandprecipitation.
Additional features in spatial variation in the mois-
tureprofileofimportancetoprecipitationareevident
from comparisons of Figs. 6a–c. In the western cross
section (Fig. 6c), the enhanced low-level moisture
covers a much greater extent into the higher latitudes
than in the eastern cross section (Fig. 6a). Positive

FIG. 4. JJA (a) 10-yr mean precipitation and (b) cold phase
anomaliesfortheAMO-forcedWRFsimulations.(c)The1971–
2000precipitationclimatology.ThesamescalesareusedasinFig.2.
Unitsareinmillimetersperday.For(b),thehatchedpatternin-
dicatesstatisticalsignificanceatthe95%confidencelevel.
284 JOURNAL OF CLIMATE VOLUME26
FIG.5.JJAmeanmoisturedivergencefor(a)coldand(b)warmphaseoftheAMO.Positivesaredivergence.Unitsarein1026kgs21m22.
moisture anomalies also extend farther into the mid- Such process may be identified by examining the at-
troposphere.Thesefeaturesindicatemoremoistening mospheric water budget given as d(rq)/dt5E2P2
in the central Great Plains in the cold phase of the $(cid:2)(rqV).Inthisbudget,Eisthesurfaceevaporation,P
AMO.Comparisonsofthethreecrosssectionsfurther istheprecipitation,$(cid:2)(rqV) isthehorizontalmoisture
reveal that the vertical gradient of moisture is more divergence, r is air density, q is the vapor mixing ratio,
prevalent in the eastern and western cross sections and V is horizontal velocity. According to the atmo-
and less prevalent in the central cross section. In ac- spheric water budget, when there is little variation in
cordance with these moisture anomalies across these moisture flux convergence between two environments,
regions, strong positive precipitation anomalies were the surface evaporation E can make the difference for
observedinNebraskaaswellasinArkansas(Fig.4b). changes in atmospheric moisture content d(rq)/dt and
Theresultsshowgoodagreementbetweenthedecreased precipitationbetweenthetwoenvironments.Figure7a
moist static stability and increased precipitation during showstheJJAanomaliesofsurfaceevaporationinthe
theAMOcoldphaseandsuggestthattheAMO-forcing coldphaseoftheAMO.Indeed,positiveanomaliesare
produces an environment that is less stable for con- shown across the central United States where an in-
vection and more efficient in converting the moisture creaseinlowertropospheremoisturewasshowninFig.5.
intoprecipitation. Particularly, the strongest positive E anomaly in the
Thesourceofthedrierairaloftinthecrosssectionsin centralUnitedStatesisnearArkansaswithasecondary
Fig. 6 is most likely the process presented in Hu et al. maximum in Nebraska. The consistent increase in sur-
(2011). They suggest that during the AMO cold phase face evaporation and moistening in low-level atmo-
themaritimeairmassfromtheGulfofMexicointrudes sphereinthecoldphaseoftheAMOintheabsenceof
northward and produces a frontal zone with the pre- a strong difference in low-level moisture flux conver-
existingcontinentalairmasstothenorth.Whilethismay gence between the cold and warm phase of the AMO
reasonably explain the drier mid- and upper-level air, (Fig. 5) suggests a strong positive feedback of surface
what may be causing the moistening in the lower tro- evaporation and precipitation in the cold phase of the
posphere? According to Fig. 5, the low-level moisture AMO. In this feedback, the moistening in the lower
flux convergence from the southerly flow is similar be- troposphere and drying aloft and their induced weak-
tweenthecoldandwarmphaseoftheAMO.Thissimi- ening in static stability of the atmosphere contribute
larity suggests some additional moistening process to convection and precipitation development, which
initiated/sustainedbythefrontalprocessduringthecold inturnmaintainsthesoilmoistureandsurfaceevapo-
phaseoftheAMO. ration. The coupling of the surface evaporation and