Extreme weather: envisioning Ontario agriculture · 2016-04-28 · Extreme weather: envisioning Ontario agriculture Scott Mitchell1, Anna Zaytseva1, Dan MacDonald2, and Ruth Waldick1,2

Extremeweather:envisioningOntarioagriculture

ScottMitchell1,AnnaZaytseva1,DanMacDonald2,andRuthWaldick1,2

AdaptationCanada2016

(1)

(2)

Purpose->constraints• “…createanddeliverinformationaboutprospectiveclimateextremesthatwillaffectOntario’sagriculturesectorandruralcommunities.Wewilldevelopadecisionsupportmodel(DSM)tocharacterizeriskandvulnerabilitiesassociatedwithclimatechangeandextremesinagriculture,allowinguserstoplanforandmitigaterisksbyevaluatingdifferentadaptationchoices.”

• spatialscenariomodellingframework– impactsoncropsandlivestock*• map-based,field-levelmapping;expectations• datarealities:weatherstations(time),GCMresolution• temporalscalesatwhichcansaymuchaboutfutureextremeeventsarehardtotranslatetoimpactstocropsandlivestock

• useofseasonal,phenology-linkedindiceswithlinkstospecificcrops

(some)Issueswithexistinginformation• (asyou’veheard)therearelimitationsinusinglimitedweatherdata,orclimatemodelprojections,tocharacterizeextremeweather• howextremesusuallyconsidered?(climatemodelvariability)• spatial-temporalresolutionissues/discrepancies• howarethoserelevanttofarm-scale/locallevelplanning?

• someoftheoptionswe’veconsidered• GCMoutput:customdownscaling,PCICdownscaling(tostationorgrid)• pastweatherdata:everythingavailable?“cleaned”data?

• station-basedorgrid(10kmregulargridusedbyAAFC,EC)?• temporalresolution:aggregatesummaries?Dailyvariability?

• scenarios:• GCM:AR4/AR5?Allmodels?Subset?• agriculture,demographics,economic(scale)

Whyfocusonscenariosandphenologicalimpactmodelling?• everyGCMmodelrunisascenario,notaprediction

• ecosystemresponseontopofthatimpactedbyrangeofpossiblereactions/adaptationfromallecosystemcomponents,includinghumans

• GCMslackspatialandtemporaldetail,but thereisdemandforinformationrelevanttolocallyevaluatinglevelsofriskandpotentialtradeoffs• finerresolutions(space&time)à assumptions&potentiallyveryhighdataneeds• usuallycan’tconfidentlyfillallthoseneeds,butcanexplorealikelyrange,considersetsoflikelyparametersunderfuturealternatescenarios

• cropmodellingtypicallyfocusesonyield,usingeitheraprocess-basedapproach(highuncertaintyinparameterizationacrosslargeregions)orempiricalmodels(usuallyassumingstationaryconditions)• phenologicalimpactmodellingallowsustoidentifytimeswhencropsareparticularlyvulnerabletoclimatologicalevents,andassignatypicalimpacttocropyield;concentrateonrelativeimpactsratherthanspecificphysiologicalprocesses

Studyarea:easternOntario

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Indicesderivedfrom“just”weatherdata

• E.Ontarionotexpectedtobeahotspotofweatherextremes• buttypesofextremesofparticularrelevancein“regular”agriculturaloperationsarenotnecessarilywhatpeoplefirstthinkofas“extreme”

• “standard”indicesareavailabletoanalyse andcompareweather/extremes• usefultodescribegeneraltrends

• some,however,maskprocessesthatareimportanttoagriculture

Whyextremes?ThisisNOTthewholestory!

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Example:generalindexrelevanttohumanhealth• 3dayperiodswhereTmax >32°C

Example:extremeindex:warmnights

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Example:precipitation

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Example:precipitation

CUMULATIVEDRYDAYS VERYWETDAYS

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

MORECROPRELEVANT:SEASONALPHENOLOGYINDICES• Corn(forexample):

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Example:poorseedingconditions

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Example:earlyflooding

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Example:seeddevelopmentdrought

A.Zaytseva’s DRAFTM.Sc.Thesis(CarletonUniversity).

Example:projectedseedingdelays

0

2000

4000

6000

8000

10000

12000

2011 2012 2013 2014 2015 2016 2019 2020

Area"lost"tofallowduetoseedingdelays

541 543 544 545 546 547

Lessonsandfutureconsiderations:

• crop- andphenology-specific,scenarioimpact-basedapproachtoextremesallowsustohighlightrelativerisksof“subtle”butagriculturallyrelevantshiftsinclimate• relevance:impactonfarmoperations• potentialtoevaluateswitchingto(orneedtodevelop)differentvarieties

• scenariomodelling:usesfield-leveldecisionsbutdoesnotrelyonneedingtoconfidentlyparameterizefield-leveldetailswithaspecific“reality”• relevanttocategoriesoffarmingoperationsastheyexistinthisregion,withrealbiophysicalconstraints• allowsustomanageuncertainty,andconcentrateonscenariosthathaverelevancetoadaptationplanning