ApplicationLaboratory InnovationsfromEarthSciences NewFrontiersinClimatePredictions: DevelopingClimateServicesforaSustainable World SwadhinBeheraandToshioYamagata Modes of Climate VAIRATION ElNino Modoki NingalooNino JAMSTEC seasonal prediction system based on SINTEXͲF1 SINTEX-F semi-multi-model ensemble system N Non-flux fl adjustment dj t t 2.2 ATMOSPHERE: ECHAM4 T106L19 T106 L19 5 OCEAN: OPA8.2 ORCAR2 Grid 20X1.50 Eq-0.5 Level 31 9-member ensemble hindcast experiments 1.Three models with different coupling physics: (Each model has realistic ENSO & IOD simulations, Luo et al. 2005a) m1: Ocean surface is solid to atmosphere. (|Ua| Ua for Tau & heat flux) m2: Ocean surface current momentum is passed to atmosphere. (|Ua-Uo| (UaUo) for Tau & heat flux) m3: Ocean surface is solid to atmosphere atmosphere, but (|Ua (|Ua-Uo| Uo| (Ua-Uo) (Ua Uo) for Tau) 2. Three initial conditions for each model: • Model spin-up (1971-1981) • A simple i l coupled l d SST SST-nudging d i iinitialization iti li ti scheme h • 䠍day, 2 days, 3 days (weekly NCEP Reynolds data) Forecast: 12 months from 1st day of every month. SINTEX-F Prediction Skills http://www.jamstec.go.jp/frcgc/research/d1/iod/index.html Basedona9ͲmembersemiͲmultimodelensemble ENSOPredictionSkillsin10CoupledGCMs predictionsystem di ti t Nino3.4 index >0.9 (1982-2001) Adapted from Jin et al. 2008, APCC CliPAS IndianOceanDipolePredictability EIO(1982Ͳ2004) Both winter and spring barrier exist Luo et al., J. Climate, 2005,. Luoetal.,J.Climate,2007,2178Ͳ2190. MMEpredictionskillsat1Ͳmonthleadtime 2m Temp Precipitation MMEskillsof2 MME skills of 2Ͳm mairtemperatureandprecipitationforJune air temperature and precipitation for JuneͲAugust August(leftpanels)and (left panels) and DecemberͲFebruary(rightpanels)for1981–2003obtainedfrom14CliPASmodels’(Wang etal.2009) ScalesofWeatherandClimate Weather Climate Variations Climate Change El Ni /IOD ElNino/IOD MJO/ISO Decades to Centuries Cyclone Seasons Tornado Whirlwind Weeks Days Hours Minutes NewChallengesinClimatePredictions •CapturinglandͲsurfaceandothernonͲlinearprocesses,besidesthe correctscaleinteractionsincoupledgeneralcirculationmodelsstill remainsachallenge.Thisaffectsthemodelpredictabilityofsomeof g p y themajormodesofclimatevariationssuchasthemonsoons. •PhaseandamplitudeofintraͲseasonaloscillationsandinteractions with monsoon and other climate modes withmonsoonandotherclimatemodes •DecadalENSO,ENSOModokipredictionsandteleconnections •DecadalchangeintheIndianOceananditsinfluenceonIOD predictionsandIODteleconnection. d d l •PredictionsofsubtropicalandmidͲlatitudeclimatemodesand regionalclimatevariationssuchasBenguelaNino,NingalooNino…… g g g Temporalanomaly correlationbetween the observed JJAS theobservedJJAS rainfallandthe ensemblemeansof theindividual models.Correlations >0.2areshaded withnegative correlations < Ͳ0 2 correlations<Ͳ0.2 hatchedorstippled. Kulkarnietal. 2012 SINTEXͲFSST&RainAnomaliesPredictions SST SST Rai n Rai n SST Rai n LowͲfrequencyvariabilityinIODaffectingitspredictability NewtypeofIOD(AbortedIOD) CoralrecordsofIODvariability Nakamuraetal.2009 Predictionforthedecayphaseofnormal IOD events IODevents Doietal.2013 Predictionforthedecayphaseof aborted IOD events abortedIODevents TheDecadalChangesinTropicalPacificandElNinoModoki Composite of surface temperature for El Nino Modoki and rainfall composites for El Nino Modoki and El Nino ENSO Modoki and decadal sea level variations December 2004, El Nino Modoki Weng,,H.,Ashok,K.,S.K.Behera,S.A.Rao,T.Yamagata,2007:Impactsof RecentElNiñoModokionDroughts/FloodsinthePacificRimduringBoreal Summer,ClimateDyn., 29,113Ͳ129 December 1997, El Nino + pIOD Behera,S.,andT.Yamagata,2010:ImprintoftheElNiñoModokion decadalsealevelchanges,Geophys.Res.Lett.,37,L23702, doi:10.1029/2010GL045936. ThePredictionSkillsofSubtropicalDipoleModesinIndianandAtlanticOceans SouthAtlanticSubtropicalDipole IndianOceanSubtropicalDipole PredictabilityofNingalooNino/Nina Oceancurrentanomaliesaveragedaboveadepth of300m(vector;msͲ1)and479seasurfaceheight anomalies(shaded;cm)inFebruary2011for(a)the SINTEXͲF1 480 prediction initialized on June 1st SINTEXͲF1480predictioninitializedonJune1st, 2010and(b)theassimilationdataofGODAS. Doi,BeheraandYamagata,2013, D i B h dY t 2013 NatureResearchReports Threemonthsaveragedtimeseriesalongafixedstart timeoftheNingalooNiño Index(NNI:SSTanomaliesaveragedin108ºͲ116ºE,28ºͲ 22ºS) 22 S)for3,6,and9monthsleadpredictionbythe for 3, 6, and 9 months lead prediction by the SINTEXͲF1andobservationaldataofNOAAOISSTv2(ºC). •SINTEXͲFhasgoodskillinpredictingmostofsignificant eventsatleastoneseasonahead •However,theonsetstageandtheamplitudeoftheevents arenotwellpredicted TheLocalAirͲSeaInteractionsforNingalooNino/Nina Kataoka,Tozuka,BeheraandYamagata2013 InfluenceofNingalooNino/NinaontherainfallvariabilityofAustraliawithandwithout ENSO SensitivitytoTropicalandExtraͲtropicalSSTAnomalies Climate Services for Sustainable Societies Observation– ProcessStudies– Modeling– g Predictions– UserinteractionsͲ Capacity p y Observations Model Predictions Internationalinitiativeskeepingsocialscienceandstakeholdersinmind: WMO/WCR P/CLIVAR Initiatives Future Earth Initiatives Belmont Forum Initiatives. International Initiatives such as suchas SATREPS.. Other regional Initiatives: JPI,APN, SASSCAL… SynergyamongInitiatives/Frameworkswouldhelpsustaininginternationaleffortsandoutcomes SATREPS Project for a Disaster Early Warning System in Southern Africa JAMSTEC, U. Tokyo, ACCESS, U. Cape Town, CSIR, U. Pretoria, ARC, SAWS and SAEON Improving CGCMs Evaluating Predictability Global-scale climate prediction di ti Regional climate prediction Capacity of seasonal climate prediction in South Africa is enhanced so that it can be applied to management of environmental problems in the Southern African Region. Improving Early Warning System MajorResearchOutcomes 9More than 40 p published p papers p ((2 in Nature,, 2 in Nature Geoscience,, 2 in Nature Climate Change and 1 in PNAS) 9First successful demonstration of mid-latitude climate prediction. 9Establishment of a multi multi-model model ensemble system in South Africa. Africa 9Application of SINTEX-F Forecasts for predicting crop yield, stream flows and heat waves. 9Successful downscaling of SINTEX-F predictions for southern Africa. Beal,L.,W.P.M.DeRuijter,A.Biastoch,R.Zahn,andSCOR/WCRP/IAPSOWorkingGroup136 (M.Cronin,J.Hermes,J.Lutjeharms,G.Quartly,T.Tozuka,S.BakerͲYeboah,T.Bornman,P. Cipollini H Dijkstra I Hall W Park F Peeters P Penven H Ridderinkhof J Zinke) 2011 : On Cipollini,H.Dijkstra,I.Hall,W.Park,F.Peeters,P.Penven,H.Ridderinkhof,J.Zinke),2011:On theroleoftheAgulhassysteminoceancirculationandclimate.Nature,472,429Ͳ436. Luo,J.ͲJ.,2011:Oceandynamicsnotrequired?Nature,477,544Ͳ546. Richter,I.,S.K.Behera,Y.Masumoto,B.Taguchi,H.Sasaki,&T.Yamagata:Adifferentmechanism toexplainthedevelopmentofAtlanticNinos.NatureGeosci. 6,43–47(2013) Lübbecke2013 BjörnC.Backeberg,PierrickPenvenandMathieuRouault,2012:ImpactofintensifiedIndian OceanwindsonmesoscalevariabilityintheAgulhassystem,NatureClimateChange2,608– 612. Prediction SSTA SuccessfulSeasonalClimatePredictions 2010Ͳ11LaNina RainfallAnom Observation SSTA Summer Fall Floods in South Africa FloodsinSouthAfrica RainfallAnom CroppredictionswiththeSINTEXͲ F Riverflowpredictionswiththe SINTEXͲF ThedominantclimaticfactorsaffectingtheyearͲtoͲyearvariationsintheyieldsofmaize,soybean,riceand wheat.Thepiediagramsindicatethepercentagesofproductionthataresensitivetotemperature(orange) andsoilmoisturecontent(turquoise)aswellasthoseforwhichnohindcastswereavailable(grey)in2000. ThedarkgreyareaindicatesnonͲcropland. g y p Iizumietal.2013,NatureClimateChange TheLegacyoftheSouthAfrican SATREPS Project SATREPSProject Anewprojectisunderdevelopmenttoestablisha climateͲbasedearlywarningsystemfortheinfectious diseasesinsouthernAfrica. COMMONCOMMUNICABLEDISEASES,THEIRDISTRIBUTION,EPIDEMICPOTENTIALANDSENSITIVITYTOCLIMATE 61966 Including diarrhoeal diseases AconceptualclimateͲbasedearlywarningsystemforMalaria Epidemic 㸦㸯㸧Infectionforecastsofar Hashizume,Minakawaetal.PNAS Hashizume Minakawa et al PNAS (2009)1857–1862 㸮ࣨ᭶ The lead is too short particularly for the developing country 㸦㸰㸧Now,theadditionofclimate predictionofpossibleSouthern Africa 2ࣨ᭶ 㸦㸱㸧IfyoucanlongͲtermpredictionby increasingtheaccuracyfurther 㸲ࣨ᭶ Well in advance, it is possible arrangement of corresponding human and stockpiling of medicine. AnewSATREPSprojecttoestablishaclimateͲbasedearlywarning systemfortheinfectiousdiseases JAMSTEC-CSIR-SAWS ACCESS & JAMSTEC Weather and climate information Regional climate model (southern Africa) JAMSTECCSIR-SAWS Down scaling Weather data NEKKEN-MRC & WCU District level climate model Refinement Refinement MRC Disease transmission model Disease ill surveillance systems ALL Disease prediction model ACCESS: Applied Center for Climate & Earth System Science JAMSTEC: Japan Agency for MarineEarth Science and Technology gy MRC: South African Medical Research Council NEKKEN: Institute of Tropical Medicine, Nagasaki University Earlyy Warning g systems New coupled GCM SINTEX-F2 AGCM OGCM Coupling Sea Ice SINTEX-F1 ECHAM4.6 T106L19 OPA8.2 2×2 L31 Every 2 hour No flux correction No SINTEX-F2 ECHAM5 T106L31 OPA9 0.5×0.5 L31 Same as F1 Yes Annual mean SST bias from NOAA OISSTv2 Horizontal distribution of the annual mean of TC genesis frequency from SINTEX-F2, SINTEX-F1 and Observations Fig 10 Horizontal distribution of the annual mean of TC genesis frequency from (a) SINTEXF2, (b) SINTEX-F1, and (c) observation. Dynamicaldownscalingofglobalclimatepredictions Ratna(2013)PerformanceassessmentofthreeconvectiveparameterisationschemesinWRFfor downscaling summer rainfall over South Africa, Climate Dynamics downscalingsummerrainfalloverSouthAfrica,ClimateDynamics InterannualvariabilityintheWRFmodelsimulated precipitationanomalies Summary GlobalpatternsoftemperatureandprecipitationanomaliesarepredictedwellbySINTEXͲF1atleastat3Ͳ monthsleadtime.ThepredictabilityremainshighforseveralseasonsleadtimeinthetropicalPacificthoughit th l d ti Th di t bilit i hi h f l l d ti i th t i l P ifi th h it deteriorateselsewhere. ENSOispredictedonanextendedrangeandthepredictabilityisprettyhighupto2Ͳyrleadtime. IODiswellͲpredictedatleastaseasonahead. However,tropicalteleconnectionsaresometimenotwellͲrepresentedinthepredictions. Decadaltrendsandsubsurfaceoceanheatcontentsarefoundtoaffectthepredictability. Increaseinthemodelspatialresolution(inSINTEXͲF2)hashelpedtoovercomesomeofthemodelbiasesseen intheresultsofSINTEXͲF1. SouthernAnnularModeseemstoplayaroleinthesubtropicalclimatevariationsandperhapsindirectlythe tropicalandotherclimatevariations NewregionalmodesneedfineͲscaleclimatepredictionsandwouldbethenewfrontiersinclimatepredictions besidesthechallengesofreachingthepotentialpredictabilityofglobalandregionalmodes.