Title: Ensemble simulations of the Last Millennium using an Earth System Model including the carbon cycle
1Ensemble simulations of the Last Millennium using
an Earth System Model including the carbon cycle
- Johann Jungclaus
- and the Millennium Team
Max-Planck-Institut für MeteorologieKlimaCampus,
Hamburg
2For the first time, millennium simulations are
carried out in ensemble mode using a
comprehensive Earth System Model (ESM) including
an interactive carbon cycle
3- Millennium was made possible by
- significant progress in the field of the
terrestrial carbon cycle (JSBACH) and land cover
change - New developments in treating volcanic forcing and
new natural forcing data sets - technical developments in model performance, and
data post-processing and storage - Computational resources as consortial experiments
at DKRZ
4The last Millennium climate
IPCC, 2007
- To what extent are the observed pre-industrial
climate variations driven by natural forcings
(orbital, solar, volcanic)? - How did the climate system respond to human
activities (land use changes, industrialization)? - What are the relations between forcing, climatic
states, and variability patterns (AO, ENSO, SAM)?
5The last Millennium CO2
Gerber et al., 2003
- Can we simulate a stable carbon cycle over the
millennial time scale with a complex Earth System
Model (ESM)? - How did the carbon cycle respond to natural and
anthropogenic disturbances and how important are
carbon-climate feedbacks?
6The Earth System Model
7External forcings
Volcanoes Time dependent information (T.
Crowley, Univ. Edinburgh) on four latitude bands
in terms of aerosol optical depth and effective
radius distribution (talk S. Lorenz)
Land Cover Changes Reconstruction of
anthropogenic land cover changes for the last
millennium (PhD thesis J. Pongratz) Fractional
maps for land use types crop, C3 and C4 pasture
at 0.5resolution combined with potential
vegetation map
8External forcings solar (TSI)
- Recent reconsiderations have drastically reduced
the range of solar variations from the Maunder
Minimum to present - The larger amplitude secular irradiance changes
of the initial reconstruction are likely upper
limits of long-term solar irradiation
variability (Lean, PAGES Newsletter, 2005)
9List of experiments
- long (3000yr) control experiment under year 800
conditions - ensemble of five simulations with all external
forcings (800-2005) - ensemble of three with alternative solar forcing
- Extension of (2) and (3) to the future (IPCC A1B)
- Single forcing experiments (one each)
- Sensitivity experiments (e.g. different scaling
for forcing components) - Pilot studies with higher resolution
(stratosphere) atmosphere model
gt15000 simulated years available through CERA
data base at the WDCC Hamburg
10Simulation of CO2 evolution
control
full forcing ensemble mean
11Simulation of CO2 evolution
Mauna Loa
full forcing expt. (monthly means)
12Observations of CO2 evolution
Ahn, pers. comm., 2009
13Simulation of CO2 evolution
Solid lines full forcing ensemble E1 (Krivova
solar, 0.1)
dashed lines full forcing ensemble E2 (Bard
solar, 0.25)
Grey shading Overlap of reconstructions (C.
Reick)
14Competing effects of individual forcings
Atmospheric CO2 concentration
control
Volcanoes only
Land-cover-change only
Global surface air temperature
15Simulation of CO2 evolution
control
full forcing (annual means)
- External forcings affect the carbon cycle
significantly volcanic eruptions lead to an
decrease, land-use-change to an increase in atm.
CO2 - How is the carbon redistributed?
16Effects of volcanoes on carbon storage
- atmosphere looses carbon mainly to the land pools
- Higher uptake by tropical vegetation and soil
17Effects of volcanoes on carbon storage
Total land carbon inventory Mol (C) m-2
year 1267 minus year 1257, ensemble means
18Northern Hemisphere temperaturesthe
instrumental period
- Simulation captures warming trend over 19th/20th
century - Observed multidecadal variations partly due to
internal variability
19Northern Hemisphere temperaturesthe last 1200
years
Background shading overlay of reconstructions
(after IPCC, 2007)
solid 5 full forcing expts. (Krivova solar
0.1) dashed 3 full forcing expts. (Bard solar
0.25)
- Range of variability consistent with
observations, but LIA cooling less pronounced
than in reconstruction for 0.1
20NH (land) temperaturesexperiments mil0012 and
mil0015 differences
Temp. Diff (K)
21Atlantic Meridional Overturningexperiments
mil0012 and mil0015 differences
AMOC (Sv)
Internal variability on multi-centennial may be
important for climate evolution, such as the
MCA-LIA transition Ensembles are essential!
22Sensitivity expts. on solar forcing
Sensitivity experiments strong vs. weak solar
variations (black and grey)
23Sensitivity expts. on solar forcing
Solar 0.25
control
Solar 0.1
24Sensitivity expts. on solar forcinglead/lag
regression of CO2 and T with TSI
T K/std.dev in TSI
CO2 ppm/std.dev in TSI
Lead/lag time (years)
Lead/lag time (years)
- Std. dev of (decadal) TSI variations is roughly
1Wm-2 - 1 Wm-2 TSI increase makes 0.1K warming and 0.7ppm
CO2 increase. Can this be identified in
observations?
25Millennium and COSMOS
- Millennium provides to COSMOS
- boundary conditions for regional studies
- a well-tested modelling framework (now COSMOS
1.2) for sensitivity experiments, scenarios, and
a test-bed for new developments (e.g. isotopes,
nitrogene cycle) - a data base of Millennium simulation accessible
through CERA (http//cera-www.dkrz.de/CERA) - call for analyses projects
- users are asked to submit a sort summary on the
planned investigations (similar to PCMDI)
26- Titel Interannual to Decadal Climate Variability
over the North Atlantic Region and its
Teleconnections in the Millennium Simulations - PI Irene Fischer-Bruns
- Institution Max Planck Institute for Meteorology
- Add. Investigators Davide Zanchettin, Johann
Jungclaus, Wolfgang Müller, Daniela Matei - Abstract
- Here we explore the millennium simulations with
respect to the climate variability in the North
Atlantic/European (NAE) region on interannual to
decadal timescales. For this, prominent patterns
of atmospheric circulation such as the North
Atlantic Oscillation (NAO) are considered.
Firstly these patterns are examined whether they
respond to the forcings applied in the different
Millennium simulations. Second we consider
whether the remote response of the NAE climate
variability to SST variations in the Pacific and
Atlantic is significantly altered. For this we
explore the prominent modes of climate
variability in the Pacific, namely the El
Niño/Southern Oscillation (ENSO) and Pacific
Decadal Oscillation (PDO), and how they are
linked with the NAE region. - Publications
- .
27Towards higher resolution MILLENNIUM Expt.
- Originally, we planned to run the Millennium
experiments with the IPCC AR4 set-up (ECHAM
T63L31, MPIOM GR1.5). This turned out to be too
computationaly demanding on the SX6. - The last Paleo Model Intercomparison workshop
(Colorado, Sept. 08) called for a Millennium PMIP
(and PCMIP) - WGCM has adopted the PMIP call and asks for
Millennium experiments as part of CMIP5 - Should be carried out with AR5 models or systems
close to it
28Towards higher resolution MILLENNIUM Expt.
- For CMIP5 core activities, presently a system
with ECHAM5 in T159/L95 and MPIOM at 0.4/L80 is
developed - For CMIP5 long term simulations (TIER1, TIER2),
such as Millennium, a lower resolution version is
assembled with ECHAM T63/L47 and MPIOM TP1/L40
29Millennium and COSMOS
- Direct simulation of proxies
- AWI-Bhv d18O Implemention in ECHAM5 almost
completed, planned for coupled model - Carbon isotope composition of carbon dioxide in
the ESM M. Cuntz (formerly MPI-BGC), Chr. Reick
(MPI-Met) - computes multiple fractionation processes
occuring during uptake of CO2 from the atmosphere
by the terrestrial and marine biosphere. - Stable as well as radiocarbon activities
- ? Coordination within COSMOS ?
30Millennium and COSMOS
- Effects of solar variation on stratospheric ozone
and feedbacks - Large variability in solar UV ? stratospheric
ozone ? polar night jet ? dynamical coupling down
to the troposphere - Approach (Hauke Schmidt)
- Distribute the variation in the solar constant
unproportionally to the bins of the SW heating
scheme. - Parameterize the stratospheric ozone change
according to the solar activity. - Possible extension
- Can we go in the direction of an interactive
(ozone) chemistry while being able to run, at
least, 100 year time slices (e.g., Maunder
Minimum vs. control)?
31Summary/outlook
- millennium simulations are carried out in
ensemble mode using a comprehensive Earth System
Model (ESM) including an interactive carbon cycle - Volcanic eruptions stand out as most powerful
forcing - External forcings affect the carbon cycle and
leave a long-lasting imprint through changes in
carbon storage - Ensembles are essential for the interpretation of
observed variability - Next steps analyses of data set in collaboration
- Higher resolution model (T63/L47 atmosphere, 1
ocean) - Participate in PMIP3/PCMIP/CMIP5
Thank you for your attention!
32(No Transcript)
33External forcings volcanoes
Pinatubo
Tambora
Huaynaputina
Krakatau
Time dependent information (T. Crowley, Univ.
Edinburgh) on four latitude bands in terms of
aerosol optical depth and effective radius
distribution
34External forcings land cover changes
cropland distribution
- Reconstruction of anthropogenic land cover
changes for the last millennium (PhD thesis J.
Pongratz) - Fractional maps for land use types crop, C3 and
C4 pasture at 0.5resolution combined with
potential vegetation map - Based on existing data (1700-1992) and population
density in earlier times