flexpart forward and backward and backward trajectories...
TRANSCRIPT
FLEXPARTforwardandbackwardtrajectoriesduringIOP1andIOP2basedon625mresoluConwindsfromWRF
simulaCons
Presented by Henrique Barbosa, Universidade de São Paulo,
on behalf of many partners of GoAmazon2014/15
GoAmazon2014/15WorkshopHarvard,May20th
on700mandIOP2andbackward
WRFandFlexpart
Goal:• UsewindsfromahighresoluConWRFsimulaConsimulaContoforceFLEXWRFLagrangianTrajectoryModel
Deliveries:• Backwardandforwardtrajectories(clusters);• Times’plumeisforecastedovereachsite;
WRF3.6.1• Period:27jan2014to1apr2014• IniCal&Boundarycond.– ClimateForecastSystemReanalysis–CFSR/NCAR– 0.5x0.5deg,every6h
• 4Grids(700m,2.1km,6.3km,18.9km)• ParameterizaCons:– Microphysics:Thompson– RadiaCon:RRTMG– PBL:Mellor-Yamada-Janjic– Land-Surface:unifiedNoah– Cumulus:Kain-Fritsch(notfor700m)
• Outputevery30min
Grids
• Grid118.9km
• Grid26.3km
• Grid32.1km
• Grid4700m
ALBEDO• VegetaCon:PROVEG-INPE(ThankstoM.Bela)
WRFINPE
GreennessfracCon• VegetaCon:PROVEG-INPE(ThankstoM.Bela)
WRFINPE
Sohowwelldidthemodeldo?
IOP1,averageoverT3site
1)Howtovalidateagainstmeasurements?
Flexwrf3.2(basedonPILT/flexpart6.2)
• ConCnuousforwardrun– SingleconCnuousrun(IOP1)– Outputevery30min– Regulargrid(@700m)and14verCcallevels
• 50m,100,200,300,600,1km,1.21.5,2,3,4,5,6,10
– ParCclereleases• 2000passivetracersevery30min(nodielcycle)• 50mto150maboveManausbox
CPC#COUNTS,"GoAmazon2014/5,"IOP1,"17"March"2014,"16:24"to"17:31"UTC!
Slide prepared by Scot Martin
Slide prepared by Scot Martin
Scot’stalk
GoodagreementwithG1measurements(for
thiscase!)
17March201416:24to17:31
UTC
Ground“concentraCon”
Lengthofplumeevents
Time(h)Insideplume
T3 2.9±2.8T2 4.3±4.3T1 ---T0e 6.1±4.7T0z 4.0±3.6T0a 0.5
2)Isthatwhatweget@groundlevel?
Length(hs)ofin-plumeevents
Ground“ConcentraCon”Average Plume>0 >1%MAO >10%MAO
T3 0.014 0.06622%
0.1014%
0.283.7%
T2 0.24 0.4850%
0.6239%
0.9525%
T1 1 1100%
1100%
1100%
T0-embp 0.44 1.432%
1.627%
1.922%
T0-zf2 0.056 0.3815%
0.511%
0.806.7%
T0-aEo 0.000044 0.0120.34%
0.0220.13%
------
AverageconcentraConsatcloseup-windsitesseemstobelargerthenatdown-windsites!
3)Isorderofmagnitudesimilartoobs?
Plumegoingforwardisdiluted!
T0-embp
T0-zf2
Plumegoingbackwardsisconcentrated!
VerCcalprofileof“concentraCon”
Log10(NormalizedConcentraIon)
AverageIOP1@50m
Log10(NormalizedConcentraIon)
AverageIOP1@200m
Log10(NormalizedConcentraIon)
AverageIOP1@300m
Log10(NormalizedConcentraIon)
AverageIOP1@600m
Log10(NormalizedConcentraIon)
AverageIOP1@2km
Log10(NormalizedConcentraIon)
AverageIOP1@3km
Log10(NormalizedConcentraIon)
AverageIOP1@4km
Log10(NormalizedConcentraIon)
AverageIOP1@5km
Log10(NormalizedConcentraIon)
AverageIOP1@6km
Log10(NormalizedConcentraIon)
AverageIOP1@10km
Longitude=T2
Longitude=T3
TimeCorrelaCon:GNDxAlCtudes
200to600mnotagoodchoice!
50and100mhavegoodcorrelaCon!
Conclusions+NextSteps
• Conclusion(careful)– Plumenotsteadyatall!Δt~0-5hsatT3– N-NEpropagaConismoreloaded– ParCclesreach10km,buthighcorrelaCon<150m
• Next(important)– Howtovalidatethemodel?ComparewithG1flightalso?Shouldwesetupaworkinggroup??
– ShouldweallowparCclestodryorwetdeposit?WhatalCtudesareinteresCng?
– Whatoutputisneeded?Traj-files?Plume-Cmes?
Thankyou!