on the astronomical forcing of simple conceptual ice age ... · on the astronomical forcing of...

1
- V, normalized ice volume (arbitray units) - I, normalized summer insolation at 65°N (arbitrary units) - 2 states model : glaciation and deglaciation - i , g , d : time constants (kyr) - Glaciation state : dV/dt = -I/ i + 1/ g - Deglaciation state : dV/dt = -I/ i – V/ d Glaciation start : when insolation is low enough (I < I 0 ). Deglaciation start : when combination of insolation and ice volume reaches the deglaciation threshold V 0 (I + V > V 0 ). Ice volume model, adapted from [1] with summer insolation at 65°N as input Ø What is the influence of the summer insolation choice on the model results ? Several summer insolation definitions : - Insolation at the summer solstice (blue) - Caloric season insolation : energy received during the half of the year with the greatest insolation intensity (brown) - Integrated summer insolation (ISI) above a given threshold [2] : - Above 300 W/m 2 (violet) - Above 400 W/m 2 (orange) On the astronomical forcing of simple conceptual ice age models Gaëlle Leloup 1,2 ([email protected]), Didier Paillard 1 1 Laboratoire des Sciences du Climat et de l’Environnement (CEA-CNRS-UVSQ ), 2 ANDRA Fig 1b) : Spectral analysis Fig 1a) : Normalized summer insolation Obliquity Precession « 100 kyr world » Larger ice sheets « 41 kyr world » Smaller ice sheets 1 000 kyr BP 1 800 kyr BP Time (kyr BP) 5 : Optimal threshold V 0 Only the solstice insolation model works for the whole Myr with a constant V 0 . For the 3 other summer insolation, the last deglaciation occurs one insolation peak too early (at MIS 3, around 50 kyr BP) but would occur at the right time with an increased deglaciation threshold V 0 for the last transition. The optimal V 0 is lower than for the 0 – 1 Myr BP period for each insolation. Model/ data agreement is not as good as for the 0 – 1 Myr BP period. No clear trend (increase or decrease) of the optimal deglaciation threshold V 0 between the 1 – 1.8 and 1.8 – 2.6 Myr BP periods. Ice age conceptual model Summer insolation Model results Data comparison Conclusions References Power [1] Amplitude and phase of glacial cycles from a conceptual model, Parrenin Paillard, 2003, EPSL. [2] Early Pleistocene Glacial Cycles and the Integrated Summer Insolation Forcing, Huybers et al., 2006, Science [3] A simple rule to determine which insolation cycles lead to interglacials, Tzedakis et al., 2017, Nature Fig 2c) : Model results for 1.8 – 2.6 Myr BP 4.5 4 5 4.8 3 2.5 2.5 3.3 3 3.5 2.5 2.5 Similar curves but different contributions from obliquity and precession The deglaciation threshold V 0 is held constant over each time period and choosen to fit best the data (optimal threshold V 0 ). Model results and data are normalized on each time period. Ice ages can be well represented by a 2 state - glaciation and deglaciation - conceptual model. Most of the transitions are modelled correctly, but the ice volume around 1180, 1830 and 2030 kyr BP periods is not, irrespective of the insolation used. Some transitions are more sensitive to the input insolation (I) and the deglaciation threshold V 0 than others. Larger glaciations (« 100 kyr world ») correspond to higher deglaciation threshold V 0 than smaller glaciations (« 41 kyr world »). Fig 3) optimal deglaciationthreshold V 0 for each time period Perspectives and open questions Ø Which physical mechanisms could explain the changes over time of the deglaciation threshold V 0 ? Ø How sensitive to parameter changes is the timing of deglaciations ? And what are the most frequent alternative deglaciations with different parameters ? Results suggest that the last deglaciation corresponds to a further increase in V 0 , to avoid a too early deglaciation at MIS 3. As found in [3], a deglaciation at MIS 3 (around 50 kyr BP) instead of MIS 1 is a frequent alternative. Fig 2a) : Model results for 0-1Myr BP Fig 2b) : Model results for 1- 1.8 Myr BP Normalized ice volume Normalized ice volume Normalized ice volume i , g , d are set to 9, 30, 12 kyr For each insolation and time period the deglaciation threshold V 0 is varried. Results for the V 0 value which best fits the data are displayed (optimal V 0 ). V 0 V 0 V 0 1180 kyr BP 1830 kyr BP 2030 kyr BP MIS 3

Upload: others

Post on 25-Sep-2020

4 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: On the astronomical forcing of simple conceptual ice age ... · On the astronomical forcing of simple conceptual ice age models Gaëlle Leloup1,2 (gaelle.leloup@lsce.ipsl.fr), Didier

- V,normalized ice volume(arbitray units)

- I,normalized summer insolationat65°N(arbitrary units)

- 2statesmodel:glaciationanddeglaciation

- 𝜏i,𝜏g, 𝜏d:timeconstants(kyr)

- Glaciationstate :dV/dt =-I/𝜏i +1/𝜏g- Deglaciation state:dV/dt =-I/ 𝜏i – V/ 𝜏d

• Glaciationstart :when insolationis lowenough (I<I0).

• Deglaciation start :when combination ofinsolationandice volumereaches thedeglaciation threshold V0 (I+V>V0).

• Ice volumemodel,adapted from [1] with summer insolationat65°NasinputØWhat is theinfluenceofthesummer insolationchoice onthemodelresults ?

Several summer insolationdefinitions :

- Insolationatthesummer solstice(blue)

- Caloric season insolation:energy received duringthehalf oftheyear with thegreatest insolation

intensity (brown)

- Integratedsummer insolation(ISI)above agiventhreshold [2]:

- Above 300W/m2 (violet)- Above 400W/m2 (orange)

On the astronomical forcing of simple conceptual ice age modelsGaëlleLeloup1,2 ([email protected]),DidierPaillard1

1LaboratoiredesSciencesduClimatetdel’Environnement(CEA-CNRS-UVSQ),2ANDRA

Fig 1b):SpectralanalysisFig 1a):Normalized summer insolation

Obliquity Precession

« 100kyr world »Larger ice sheets

« 41kyr world »Smaller ice sheets1000kyr BP 1800kyr BP Time(kyr BP)

5 :Optimalthreshold V0

Only thesolsticeinsolationmodelworks forthewhole Myr with aconstantV0.Forthe3other summer insolation,thelastdeglaciation occurs oneinsolationpeak too early (atMIS3,around 50kyr BP)butwould occur attherighttimewith anincreased deglaciation threshold V0 forthelasttransition.

TheoptimalV0 is lower than forthe0– 1Myr BPperiod foreach insolation.Model/dataagreementis notasgoodasforthe0– 1Myr BPperiod.Noclear trend(increase ordecrease)oftheoptimaldeglaciation threshold V0between the1– 1.8and1.8– 2.6Myr BPperiods.

Ice age conceptual model Summer insolation

ModelresultsDatacomparison

Conclusions References

Power

[1]Amplitudeandphaseofglacialcyclesfromaconceptualmodel,Parrenin Paillard,2003,EPSL.[2]EarlyPleistoceneGlacialCyclesandtheIntegratedSummerInsolationForcing,Huybers etal.,2006,Science[3]Asimpleruletodeterminewhichinsolationcyclesleadtointerglacials,Tzedakis etal.,2017,Nature

Fig 2c):Modelresults for1.8– 2.6Myr BP

4.5

4

5

4.8

3

2.5

2.5

3.3

3

3.5

2.5

2.5

Similar curves butdifferent contributionsfromobliquity andprecession

Thedeglaciationthreshold V0 is heldconstantovereachtimeperiod andchoosen tofitbestthedata(optimalthresholdV0).Modelresults anddataarenormalized oneachtimeperiod.

Ice ages can be well represented bya2state- glaciationanddeglaciation -conceptual model.

Mostofthetransitionsaremodelled correctly,buttheice volumearound1180,1830and2030kyr BPperiods is not,irrespective oftheinsolationused.

Some transitionsaremoresensitivetotheinputinsolation(I)andthedeglaciation threshold V0 than others.

Larger glaciations(« 100kyr world »)correspondtohigher deglaciationthreshold V0 than smaller glaciations(« 41kyr world »). Fig 3)optimaldeglaciationthresholdV0

foreach timeperiod

PerspectivesandopenquestionsØ Which physical mechanisms could explain thechangesover

timeofthedeglaciation threshold V0 ?

Ø Howsensitivetoparameter changesis thetimingofdeglaciations ?Andwhat arethemost frequentalternativedeglaciations with different parameters ?

Results suggest that thelastdeglaciation correspondstoafurther increase inV0,toavoid atoo early deglaciation atMIS3.

Asfound in[3],adeglaciation atMIS3(around 50kyr BP)instead ofMIS1is afrequent alternative.

Fig 2a) :Modelresults for0-1MyrBP Fig 2b) :Modelresults for1- 1.8Myr BP

Norm

alize

dicevolume

Norm

alize

dicevolume

Norm

alize

dicevolume

𝜏i, 𝜏g, 𝜏daresetto9,30,12kyrForeach insolationandtimeperiod thedeglaciation threshold V0 is varried.Results fortheV0 valuewhich bestfits thedataaredisplayed (optimalV0).

V0 V0 V01180kyr BP 1830kyr BP 2030kyr BP

MIS3