Upload
others
View
6
Download
0
Embed Size (px)
Citation preview
DecliningSnowCoverReducesRadia4veCoolingfromHistoricLandUseChangeintheWesternGreatLakesRegionBethanyBlakely,AdrianRocha,JasonMcLachlan
DepartmentofBiologicalSciences,UniversityofNotreDame,NotreDameIN
°
LandSurfaceChanges CombinedForcing
Acknowledgements
LiteratureCited
Ques4ons
Reconstruc4ngHistoricVegeta4on
v ForestcoverhasdecreasedsinceEuropeanseMlementv Deciduous and mixed forests have mostly replacedevergreenforestswhereforestregrowthisoccurring
Figure2:Vegeta4oncoveronthe(a)historiclandscape(b)modernlandscape.Percentchangeinvegeta4oncoverisshownin(c).WeusedgriddedPublicLandSurvey(PLS)data to assign land cover classifica4ons using Interna4onal Geosphere BiosphereProgramme (IGBP) designa4ons as follows: evergreen forest ( > 60% evergreen sp.cover), deciduous forest (> 60% deciduous sp. cover), mixed forest ( > 60% forestcover,mixedcomposi4on),vegeta4onmosaic(20%-60%forestcover),andcropland/grassland(<20%forestcover).
TheMidwestintheAnthropocene
v Land use and snowcoverhavechangedin the Great LakesR e g i o n s i n c eE u r o p e a nseMlement
v These changes alterthe brightness andtemperature of theland surface withimp l i ca4ons forclimate
Figure 1: Typical land use history in theGreatLakesRegion. Landscapes (a)beforedeforesta4on, (b) immediately a_erdeforesta4on,and(c,d)today.
IwouldliketothankA.RochaandJ.McLachlanfortheirideasandguidanceonthisproject,X.YangandB.NaimifortheirMODISprocessingscripts,theMcLachlanlabundergraduatesfordigi4zingPLSdata,andS.Goringforcrea4onofthegriddedPLSproduct.ThisprojectisfundedbytheArthurJSchmiMFounda4on,thePaleonproject,andtheUniversityofNotreDame.
v How have historic changes in land usea l t e red t he a lbedo and su r f a cetemperatureoftheGreatLakesRegion?
v What are the radia2ve forcings of thesechanges and how do they offset eachother?
v Howhastheclima4callydrivendecreaseinsnowcoverimpactedtheseeffects?
Reconstruc4ngLandSurfaceTraits
SnowCoverChanges
MODISAlbedo
MODISSnow
MODISLandCover
MODISQualityControlandTen-yearAggrega4on
ReconstructedVegeta4on
MODISSurfaceTemp.
MODISLandCover
ReconstructedAlbedo ReconstructedSurfaceTemp.
Figure3:Methodologyforreconstruc4nghistoricalbedoandsurfacetemperature,calcula4ngchangessinceEuropeanseMlementandcalcula4ngradia4veforcingofthosechanges.Interpola4onsofmodernMODISdatawereusedtoassignbiophysicalproper4estothemapofhistoricallyreconstructedvegeta4on.
Albedo SurfaceTemperatureProcess
Model
Assign
AlbedoDifferencesandRadia4veForcing
SurfaceTemp.DifferencesandRadia4veForcing
Compare
• Chen,J.,Jönsson,P.,Tamura,M.,Gu,Z.,Matsushita,B.,&Eklundh,L.(2004).Asimplemethodforreconstruc4ngahigh-qualityNDVI4me-seriesdatasetbasedontheSavitzky–Golayfilter.RemoteSensingofEnvironment,91(3–4),332-344.
• Zhao,K.,&Jackson,R.B.(2014).Biophysicalforcingsofland-usechangesfrompoten4alforestryac4vi4esinnorthamerica.EcologicalMonographs,84(2),329-353.
a b
c d
a b
c
The nega2ve forcings of historic land use changecurrently provide a “discount” on regionalwarmingbut thesebenefits are likely todisappearwith2meas snow cover decreases and forest regrowthcon2nues
LowessInterpola4onsofMODISdata
GHCNSWE2000-2010
GHCNSWE1900-1910
ReconstructedAlbedo
Calculatedshi_insnowseasonality:
Hsnownorm-Msnownorm
Appliedto
Seasonallyshi_edAlbedo
Seasonalshi_differencesandradia4veforcing
v Surfacetemperatureforcingisstronglyposi4ve in summer and weaklynega4veinwinter.
v Net forcing from surface temperaturechangeweightedbydaylengthissmallandposi4ve(μ=+0.37W/m2)
-2°C
+2°C
Figure 6: Surface temperature differences for Winter (Jan 1),Spring (Apr 7), Summer (Jul 12), and Fall (Oct 16). Nega4vevalues (green) indicate areas where modern surfacetemperatureislowerthanhistoricsurfacetemperature.
Figure7:Seasonalprofileof radia4ve forcing for surface temperaturechange.Grayshadingdesignates10%and90%quan4les.
v Springsnowmeltisdelayed,increasingalbedoinFeb-May
v Spring forcings aretypically an ordero f m a g n i t u d elargerthanfall
v Net forcing fromthe shi_ in snows e a s o n a l i t y i sposi4ve but small(+0.45W/m2)
Figure5:Albedodifferences forWinter (Jan1), Spring (Apr7),Summer(Jul12),andFall(Oct16).Posi4vevalues(green)indicateareaswheremodernalbedo ishigher thanhistoricalbedo.
-0.3
+0.3
SurfaceTemperature
Albedo
Figure4: Seasonalprofileof radia4ve forcing foralbedochange.Grayshadingdesignates10%and90%quan4les.
Figure 8: Spring (a) and Fall (b) shi_s in Albedo due to changes in snowcover. Black lines represent historically reconstructed Albedo shi_ed toreflect historic snow cover. Red lines show non-shi_ed historicallyreconstructedAlbedo.
Figure9:Seasonalprofileofradia4veforcingforsurfacetemperaturechange.Grayshadingdesignates10%and90%quan4les.
Figure 10: Methodology for seasonalshi_inalbedobasedonchangesinsnowcover. Global Historic Climate Network(GHCN) snow water equivalents wereusedtocalculateseasonalshi_
v Surfacetemperaturechangesoffset23%ofyear-roundalbedoforcing
v Clima4cshi_sinsnowseasonalityoffset18%ofyear-roundalbedoforcing
v Combinedforcingis41%lowerthanvegeta4on-mediatedalbedoforcingalone
Figure11:Albedo(a)andtotal(b)radia4veforcingsofland use change in the Great Lakes Region. Dashedlinesindicatecomponentforcings.
Conclusions
v Albedo forcing is always nega4vewiththegreatestcoolinginwinter
v Overall radia4ve forcing fromalbedo change is large andnega4ve(μ=-1.64W/m2)
a b
a
b