Transcription
MESA Modeling andData AssimilationMESA modeling group:I. Cavalcanti, A. Seth, C. Saulo,B. Kirtman, V. MisraCeleste Saulo – VPM9 april 2006
MESA modeling objectivesModel AssessmentModel DevelopmentHypothesis TestingCeleste Saulo – VPM9 april 2006
RESULTS OF SOMEACTIVITIESCeleste Saulo – VPM9 april 2006
Model assessmentsVerify the ability of models to simulate and predictfeatures of the SAMSIdentify model deficiencies the diurnal cycle in both regional and global modelsthe annual cycleintra-seasonal variabilityinter-annual variabilitydecadal variability (IPCC models)20th century observed climate trends (IPCC models)simulation and predictability during SALLJEXsimulation of extremes.Celeste Saulo – VPM9 april 2006
Model AssessmentDJFFrom J N-Paegle et al (VPM8, 2005)Cavalcanti 2006Menendez 2006DJF precipitation averages from R-1and 2 do not reproduce the continentalmaximum at 50-65W. This is not thecase for the COLA model. All threeestimates have spurious orographiceffects over the Andes.Celeste Saulo – VPM9 april 2006
Diurnal Cycle CMORPH dataCeleste Saulo – VPM9 april 2006
the diurnal cycle in both regional and global modelsDJFCPTEC/COLA AGCM12 GMT(1980-2001)00 GMTREGIONAL ETA18 GMT12 GMT06 GMT00 GMTCavalcanti, 2006Celeste Saulo – VPM9 april 2006(1996-2000)18 GMT06 GMT
Important discrepancies betweenmodelsIPCC AR4 AOGCMs -20 models, 65 runspresent climate (period 1979-1998)(Menendez 2005)Celeste Saulo – VPM9 april 2006
South American MonsoonPrecipitation and Moisture Fluxin the SRES A2 ScenarioMaisa Rojas (U Chile, Santiago)Anji Seth (U Connecticut, Storrs)Sara Rauscher (ICTP, Trieste)Acknowledgement: IPCC AR4 Modeling Groups and WG I for coordinating, archivingand making accessible the model integrations.Celeste Saulo – VPM9 april 2006
1970-2000Monthly PrecipitationMonsoon: models capture theannual cycle.Amazon: models simulatespurious semi-annual cycle, anddelay/underestimate observedlate summer (JFM) maximum.Southeast: modelsunderestimate summer rains(NDJF), reduce the amplitude ofthe annual cycle.Celeste Saulo – VPM9 april 2006
(2070-2100)-(1970-2000)Monthly PrecipitationMonsoon: Little agreementamong models during rainyseason (NDJFM). Drier earlyrainy season (SON), wetter laterainy season (JFM)?Amazon: Little agreementamong models during onset ofrains (SON). Most modelssuggest increased precipitationduring middle/late rainy season(DJFM).Southeast: General modelagreement towardsincreased precipitation,especially in spring (OND).Celeste Saulo – VPM9 april 2006
IPCC AR4 AOGCMs: A1B scenario projections (2079-2098)About 70% (40%) of the models project a wetter climate in australsummer and autumn (winter and spring) in AMZ, while about 50%60% of all the models project a wetter climate in SSA all over theyear. (Menendez 2005) Celeste Saulo – VPM9 april 2006
Model development Improve the seasonal prediction and weatherforecasting over South AmericaStimulate the development of physicalparameterizationsImplement data assimilationCeleste Saulo – VPM9 april 2006
Downscalingtechniques199719981999Vasubandhu Misra, 8ISCHMOJFM seasonal mean precipitation anomaliesExperimentFeatureCOLA AGCMAGCM seasonal integration atT42CONTROLCONTROL-BRSM nested into COLA AGCM inconventional mannerEXPTRSM runs with Scale SelectiveBias Correction applied onanomaly nested variables ofwinds and surface pressure.Celeste Saulo – VPM9 april 2006
DataAssimilation:a) globalMean low-level wind(vector) and wind speedmagnitude (shaded) at 850hPa for january 15 toFebruary 15, 2003. a)CDAS1, b) CDAS2, c)GDAS, d) CDAS1rp,e)CDAS2rp and f) GDASrp.Values are in m/s.Herdies et al,2006 submittedJ. of ClimateCeleste Saulo – VPM9 april 2006
Data Assimilation: b) regionalEnriched analyses weregenerated ingesting all availabledata during SALLJEX, followinga downscaling methodology,using the Regional AtmosphericModeling System (RAMS),Version BRams 3.2Skabar and Nicolini, 8SCHMOThese enriched analyses arecurrently applied to studyevolution of convection duringSALLJ events providing a muchbetter resolution of thepreconditioning processes thatgradually buildup theenvironment that promotesorganized deep convection oversubtropical South America.Celeste Saulo – VPM9 april 2006
Integration of models:Concept of Super Model Ensemble Several models are available: global, (CPTEC, NCEP, ECMWF, UKMO, ) ; Regional models in S. America: CPTEC (ETA,BRAMS), INMET(DWD regional model), MASTER (BRAMS), SIMEPAR (ARPS,BRAMS), UFRJ (MM5, WRF,RAMS), FURGS (BRAMS), EPAGRI(BRAMS), LNCC (ETA), CIMA/UBA (WRF), aprox. 14 models ! Differences in physical processes parameterization, data assimilation,data source
Bias V-comp. - at SantaCruz/BoliviaSALJJEX MeanMeridional WindComponent Bias up to7days forecastMESMES is theoptimal statisticalcombination of allavailable forecasts(near zero bias)
Evaluate scientific hypotheses tomeet MESA science objectives.The hypothesis testing should include but maynot be limited to Synergy between SALLJ and MCS,Mechanism for the NW Argentina heat low,Sensitivity to soil moisture,Coupled simulation in the region of the SACZ,Local and remote (global) influence of SAMS.Celeste Saulo – VPM9 april 2006
Land-surface impactThe majormechanisms forthe LLJ are thehorizontaltemperaturegradient causedby land surfaceheterogeneitiesand theoscillation of thefrictional effect.Wu and Raman 1997: Experiments with the NCSUmesoscale model, with idealized conditions andinitial geostrophic southerly wind constant withheight (10 m/s). Initial time 08 LSTCeleste Saulo – VPM9 april 2006
Soil moisture impactControl expZhong et al. 1996Strong impact on the amplitude.Larger amplitude with drier soilCeleste Saulo – VPM9 april 2006
The effect of soil moisture on precipitation- 30% SMControl 30% SMMonthly Averaged fieldsETA-UMD modelCollini et al. 2006Precipitation changesPrecipitation as a function of Soil moisture[20-10 S; 55-45 70-45-30-1501530SoilmoisturechangesCelesteSaulo– VPM9 april200645
The effect of soil moisture on the diurnal cycleSensible heatLatent heatBowen RatioPrecipitationBlack: ControlRed: Reduced soil moistureCeleste Saulo – VPM9 april 2006Averages over (15º S, 75º W)(5º S, 34º W). Collini et al. 2006
Several sensitivity studies employingdifferent land use and soil moisturepatterns using the WRF model wereperformed to analyze how changesin surface conditions alter not onlythe NAL but also the circulationpatterns associated with theseeventsE10,25E2E4E3Control0,2E1: DRY OBS SOIL0,15E2: VERY DRY OBS SOILE3: OBS MOIST BARE SOILE4: VERY DRY BARE SOIL0,10,050765432Celeste Saulo – VPM9 april 20061313029Ferreira et al., 8SCHMO
E2INTENSIFICATIONFerreira et al., 8SCHMOCeleste Saulo – VPM9 april 2006
MESA Modeling and Data Assimilation MESA modeling group: I. Cavalcanti, A. Seth, C. Saulo, B. Kirtman, V. Misra. Celeste Saulo - VPM9 april 2006 MESA modeling objectives . COLA AGCM AGCM seasonal integration at T42 CONTROL-B RSM nested into COLA AGCM in conventional manner EXPT RSM runs with Scale Selective Bias Correction applied on
