WGCM Chemistry

1
Effect of Chemistry in Climate Simulations and
SPARC/ACC Data Sets for CMIP5
Veronika Eyring (DLR), Ted Shepherd (Toronto),
Darryn Waugh (JHU) SPARC CCMVal Jean-François
Lamarque (NCAR), Drew Shindell (GISS) ACC4
12th Session of the JSC/CLIVAR Working Group on
Coupled Modelling (WGCM) Paris, 22-24 September
2008
2
SOLARIS Modeling and understanding the solar
influence on climate through stratospheric
chemical and dynamical processes.
CCMVal Improve understanding of CCMs through
process-oriented evaluation and provide reliable
projections of stratospheric ozone and its impact
on climate
DynVar costs and benefits of including a
realistic stratosphere in ESMs with a focus on
circulation dynamics
3
Phil Rasch, Sarah Doherty, A. R. Ravishankara
ACC4
4
OUTLINECHEMISTRY

• Effect of including a realistic stratosphere in
  climate simulations (processes and feedbacks)
• Effect of stratospheric ozone changes on climate
• Effect of climate change on ozone recovery not
  simply controlled by ODSs
• Importance of tropospheric composition in climate
  simulations
• SPARC and ACC Data sets for CMIP5

5
Ia. Effect of stratospheric ozone on SH
high-latitude climate
The ozone hole has led to changes in SH
high-latitude surface climate

• Model forced by ozone depletion

Observed (1979-2000) December-May trends
Gillett Thompson, Science, 2003
6
Trop. Jet Comparison CCMVal with AR4
modelsCCMVal models have fully interactive
stratospheric chemistry
2000-2050 trend in Zonal Wind

• Owing to the disappearance of the ozone hole in
  the first half of the 21st century (Eyring et
  al., 2007 WMO, 2007)
• Deceleration poleward side of jet (decrease in
  SAM) found in multi-CCM mean.
• Opposite response in mean of IPCC AR4
  simulations.
• Importance of ozone can be seen by comparing AR4
  models with without ozone recovery.
• Weaker response in AR4 models with O3 recovery.

CCMs
AR4
No recovery
O3 recovery
Son et al., Science, 2008 see also Perlwitz et
al., GRL, 2008
7
Testing impact of interactive chemistry
2000-2050 trend in Zonal Wind
CCM
1. GEOSCCM REF2 run. 2. GCM run with
monthly-mean zonal-mean O3 from CCM REF2 run.
Response in GCM is weaker than CCM, with
difference similar to CCM vrs AR4 with recovery.
CCMs
AR4
GCM
No recovery
O3 recovery
Courtesy of Luke Oman (JHU)
8
Testing impact of upper boundary and high
vertical resolution in the stratosphere
Surface response to the imposed ozone depletion
as measured by difference in the mean sea level
pressure from 30S to 90S
10 hPa lid momentum not conserved
10 hPa lid, cons
0.001 hPa lid
Contour interval is 1 hPa negative values are
dashed. Dark (light) shading denotes confidence
at 99 (95) level.

• The SH high-latitude surface response to the
  ozone hole is seriously compromised if the
  stratosphere is not properly represented
• Proper attention also needs to be paid to
  conservation of momentum in the gravity-wave drag
  parameterization DJ mean SLP response to ozone
  hole

SPARC DynVar Paul Kushner
Shaw, Sigmond, Shepherd Scinocca, J. Clim., 2008
9
Ib. Ozone will be affected by climate change, not
just by ODSs
Observed changes over 1979-2005, regressed
against EESC
Tropical Ozone 1980s-2040s in the CCMVal-1 models
Increase in upper stratospheric ozone associated
with CO2-induced cooling decrease in lower
stratospheric ozone associated with increased
tropical upwelling from Eyring et al., JGR, 2007
The lower stratospheric tropical ozone decrease
is also observed but it should probably be
attributed to climate change, not ODSs (and hence
is not expected to reverse) from Randel Wu,
JGR, 2007
10
The recovery of stratospheric ozone will be
affected by climate change, not just by ODSs

• Climate models consistently predict a
  strengthening of the Brewer-Dobson circulation in
  response to climate change (Butchart et al. 2006
  Clim. Dyn.)
• This leads to super-recovery of total ozone in NH
  midlatitudes, and subrecovery in the tropics, so
  that ozone evolution does not just follow
  chlorine loadings

Red is obs
Green is Cly
N.B. The obs do not contradict the model ozone
has in fact declined in the tropical lower
stratosphere
Shepherd (2008 Atmos-Ocean) see also Eyring et
al., 2007
11
Effect of including a realistic stratosphere -
Temperature trends -
1980-2004
Plot courtesy of Tesfai Cordero, San Jose State
University
12
II. Importance of tropospheric composition in
climate simulations

• Tropospheric Composition important component
  of the chemistry-climate feedback system
• Ozone
• Radiative forcing of tropospheric ozone
  (importance on Upper Troposphere/Lower
  Stratosphere, Arctic climate, surface SW flux)
• Vegetation damage gt effects on land biosphere
  CO2 uptake/storage
• OH
• reactions with SO2DMS gt sulfate and VOC gt SOA
• controls important GHG levels (esp. CH4)
• Tropospheric Oxidants needed for sulfur chemistry
• Nitrogen deposition on vegetation

13
III. SPARC and ACC Data Sets for CMIP5 (1) Time
series of OZONE based on OBS where possible and
CCM output where needed

• (I) Observed stratospheric data base built from 5
  observational databases
• The NCAR database (Randel Wu)
• The NIWA database (Bodeker Hassler)
• The NASA/NOAA/RAL database (Rosenlof Gray)
• The NASA/GSFC database (Stolarski and Frith)
• The Environment Canada database (Fioletov and
  McLinden)

led by G. Bodeker
EESC regression coefficients will be used to
extrapolate the data back in time
Extended backward with regression model (SPARC
OBS)
SPARC OBS
Extended backward in time to background ozone
levels in 1850
ACC4
1850
1900
1950
2000
2050
2100
2150
Y
e
a
r
14
III. SPARC and ACC Data Sets for CMIP5 (2) Time
series of OZONE fully based on CCM output

• Definition of emissions
• Emissions (1850-2300, every 10 years, 0.5)
  defined by an international consortium of
  historical/present-day global emission scientists
  and Integrated Assessment Modelers will be
  available by end of 2008 and will be consistent
  (1) with the 4 RCPs and (2) with the historical
  reconstruction.
• Current Status on past/future in IAMs
• try to be able to get a reasonable
  representation for 2005. Highest priority would
  be those gases that determine the climate signal
  most, i.e. CO2, CH4 ... (and would therefore be
  looked at most).
• Main base year for inventory remains 2000.
• 2005 data will be used to assess and possibly
  constrain the initial trajectory of the future
  emissions in the IAMs.

15
III. SPARC and ACC Data Sets for CMIP5 (2) Time
series of OZONE fully based on CCM output

• Definition of simulations CCMs with tropospheric
  and stratospheric chemistry to ensure continuity
  across the tropopause and consistency between
  tropospheric and stratospheric distribution (i.e.
  role of STE)
• Year 2000 to be used as internal check
• 1850-2150 (possibly 2300) to create modeled
  ozone climatology (10-year averages) in support
  of AR5 AOGCM simulations from participating ACC4
  and CCMVal models. The simulations will be
  performed mainly using fixed SSTs (from previous
  AOGCM simulations).
• Generating best ozone dataset from the
  submitted model simulations
• Submitted runs will be evaluated by ACC4
  (tropospheric ozone) and CCMVal (stratospheric
  ozone) to form a best estimate plus uncertainty

16
III. SPARC and ACC Data Sets for CMIP5 - OTHER
SPECIES AND FORCING DATA SETS -

• Additional fields that will be made available by
  ACC4
• Atmospheric oxidants
• Nitrogen deposition
• Surface ozone (or specific measure of exposure)
• Will climate models need aerosol distribution as
  could be generated by this ACC4 exercise?
• Solar Irradiance Data (SPARC SOLARIS), see
• http//www.geo.fu-berlin.de/en/met/ag/strat/resear
  ch/SOLARIS/Input_data/index.html
• Heating rates from volcanic aerosol and Surface
  Sulphate Area Densities
• can be downloaded from the CCMVal Forcing website
  at
• http//www.pa.op.dlr.de/CCMVal/Forcings/CCMVal_For
  cings_WMO2010.html

17
IV. Summary and Recommendations

• A recommendation for models that do not have
  interactive chemistry should be made to prescribe
  ozone according to the new SPARC/ACC ozone time
  series, because
• The recovery of stratospheric ozone and ozone in
  the UTLS will be significantly affected by
  climate  change important for RF
• The impact of the ozone hole on high latitude
  surface climate has  been substantial, so the
  impact of ozone  recovery in the future will also
  be important. This has implications for SH
  high-latitude climate (e.g., tropopause height,
  jet location, Hadley Cell extent, carbon uptake,
  and sea-ice melt).
• Prescribe atmospheric oxidants to capture changes
  in sulfate formation
• Prescribe nitrogen deposition to capture changes
  in nitrogen input to vegetation (if field can be
  used)
• Prescribe measure of ozone exposure (if field can
  be used)

18
V. Feedback from WGCM

• Utility of dynamically referenced databases
  Compatibility between the prescribed
  stratospheric ozone data and the internal
  dynamics of the model ?
• Databases in equivalent latitude?
• Vertical coordinate referenced to the local
  tropopause?
• Providing a measure of the variability on the
  ozone time series?
• Utility of (a) observational data base combined
  with model time series versus (b) a fully modeled
  time series?
• Advantages (a) most accurate in the 20th
  century Problems how to best combine
  troposphere and stratosphere and model output in
  the future
• Advantages (b) consistency with other species
  combined tropstrat fields Problems Tight CMIP5
  Timeline might not be accurate in the past
• Will climate models need prescribed atmospheric
  oxidants, nitrogen deposition, aerosol
  distribution ?
• Other questions
• Vertical resolution and extent
• Zonally averaged or 3D?
• Time-resolution daily, monthly, long-term trend