Institute for Climate and Atmospheric Science

1
Institute for Climate and Atmospheric
Science SCHOOL OF EARTH AND ENVIRONMENT
3D SLIMCAT Studies of Arctic Ozone Loss
Wuhu Feng
Acknowledgments Martyn Chipperfield, Stewart
Davies, L. Gunn, V.L. Harvey, C.E. Randall, M.L.
Santee, P.Ricaud, QUOBI and SCOUT-O3
2
OUTLINE

• 3D CTM SLIMCAT
• Examples of CTM results Comparison with various
  measurements
• Modelled O3 loss under different meteorology
  conditions
• Sensitivity experiments
• Conclusion

3
SLIMCAT/TOMCAT 3D CTM

• Off-line chemical transport model with many
  different options.
• Key points here
• Extends to surface using hybrid ?-? (SLIMCAT),
  ?-p (TOMCAT).
• Variable horizontal/vertical resolution.
• Horizontal winds and temperatures from (UKMO,
  ECMWF etc) analyses
• Model constrained to real meteorology--good for
  comparison with Obs. .
• Vertical motion from diagnosed heating rates
  (SLIMCAT) or divergence of mass flux (TOMCAT).
  Note analysed vertical wind can be noisy.
• Tropospheric physics convection, PBL mixing etc
• Chemistry Full stratospheric chemistry
  scheme (41 species, 160 reactions) with
  heterogeneous chemistry on liquid/solid
  aerosols/PSCs and an equilibrium denitrification
  scheme. NAT-based microphysical denitrification
  (DLAPSE) scheme included and detailed
  tropospheric chemistry scheme
• Sequential chemical data assimilation scheme
  sub-optimal Kalman Filter
• http//www.see.leeds.ac.uk/slimcat

4
Comparison with MIPAS O3 in the SH
Feng et al.(JAS, 2005)
5
Comparison with POAM O3 in the NH
Singleton et al. (ACP,2005)
6
Long-term N2O variation Comparison with satellite
and ground-based measurements
Ricaud et al.(to be submitted)
7
Effect of chemical data assimilation
Gunn et al.(to be submitted)

• Assimilation of HALOE data (ie. CH4, O3, HCl and
  H2O) into SLIMCAT reproduce better long-term NO2
  variations

8
Arctic Ozone loss versus VPSC
New T42 run
Obs
New T15 run
Old T15 run
Chipperfield et al. (GRL, 2005)

• First successful CTM simulation of seasonal O3
  column loss
• and reproduces the past climate sensitivity of
  Arctic ozone depletion on T.

9
Modeled Arctic Ozone Loss
Updated from Feng et al. (2007)

• Year-to-year variations of ozone loss due to
  different meteorological conditions
• Arctic ozone loss is initially limited by the
  availability of sunlight in early winter and
  curtailed by the breakdown on the vortex in late
  winter/spring

10
Sensitivity experiment

• More ozone loss if the Arctic winter 2004/05
  after late February was followed by 1997 and 2000
  meteorological conditions
• Arctic ozone loss would have been even more
  severe and complete loss would have occurred
  around late March if the winter 2004/05 was
  followed by 1997 conditions which had a record
  long-lasting cold polar vortex
• No Arctic ozone hole structure if the winter
  2004/05 followed by 2000 meteorological
  conditions. However, the Arctic ozone hole
  would have happened if followed by a spring like
  1997 with a long-lasting cold polar vortex.

Minimum temperature (K) at 456K from March to
April for 2005, 2000 and 1997 from ECMWF
analyses. (b) Maximum modelled local ozone loss
() at 456 K for winter 2004/05 and two
sensitivity runs where the simulation for 2004/05
was continued with meteorology for 1997 and 2000
after February 28. (c) As panel (b) but for
minimum column O3 along with TOMS data for 2005
for any point poleward of 65o N.
From Feng et al. (2007)
11
Laboratory Cl2O2 cross data
Burkholder et al. (1990) JPL (2006) Huder and
Demore (1995) Pope et al. (2007)
L

• Large discrepancy of cross section of Cl2O2 from
  laboratory measurements

12
Cl2O2 photolysis rate
Burkholder et al. (1990) JPL (2006) Huder and
Demore (1995) Pope et al. (2007)
Burkholder et al. (1990) JPL (2006) Huder and
Demore (1995) Pope et al. (2007)

• Different Cross section of Cl2O2 results in
  different photolysis rate
• JCl2O2 from recent new laboratory data is a
  factor of 6 than the current JPL recommendation
• Standard SLIMCAT CTM uses JCl2O2 values based on
  Burkholder et al.(1990) data which is the fastest
  than other data

13
Impact of absorption cross section of laboratory
Cl2O2 on the modelling ozone loss
450 K for Antarctic 2003
475 K for Arctic 2002/03
Match Burkholder et al. (1990) JPL (2006) Huder
and Demore (1995) Pope et al. (2007)
Match Burkholder et al. (1990) JPL (2006) Huder
and Demore (1995) Pope et al. (2007)

• Ozone loss rate is very sensitive to JCl2O2
  values
• Similar evolutions of diagnosed ozone loss rate
  from model runs using different absorption cross
  section of Cl2O2 from laboratory measurements in
  the polar regions
• Large discrepancy when using the new laboratory
  data (pope et al. 2007). Measurement problem or
  model still uncertain???

14
Impact of Cl2O2 cross section on the Arctic ozone
loss

• Large Arctic Ozone loss using different Cl2O2
  cross section in model

15
Comparison with MKIV data
MKIV Burkholder et al. (1990) JPL (2006) Huder
and Demore (1995) Pope et al. (2007)

• Standard SLIMCAT reproduces
• ClO very well while model using
• Pope et al. (2007) underestimated
• observed ClO

ClO
HCl
ClONO2
O3
16
Chlorine partitioning Comparison with AURA MLS

• SLIMCAT overestimate Chlorine activation? Need
  to test new PSC
• Scheme and ClOx kinetics

Santee et al..(JGR, in press)
17
Best agreement for SLIMCAT with DLAPSE
SLIMCAT overestimates chlorine activation,
importance of Liquid aerosols in CTM
18
(No Transcript)
19
Summary and Outlook

• Different measurements help testing simulations
  from CTM (ie. SLIMCAT), while the reliable model
  can be used to check the consistencies of the
  observations.
• Ozone loss is initially limited by the
  availability of sunlight in early winter and
  curtailed by the breakdown on the vortex in late
  winter/spring
• SLIMCAT reproduces the past climate sensitivity
  of Arctic ozone depletion on T
• Recent new experiment shows large discrepancy of
  cross section of Cl2O2 from other laboratory
  measurements, which results in different
  photolysis rate
• Ozone loss rate is very sensitive to JCl2O2
  values
• Standard SLIMCAT reproduces observed ozone loss
  rate quite well, while it underestimates ozone
  loss rate when using Pope et al. (2007) data.
• SLIMCAT with detailed DLAPSE microphysical
  scheme is less denitrified while model with
  equilibrium scheme has stronger denitrification,
  however, there is only small effect on ozone loss
  between these two schemes.
• Compare HIRDLS data (ie. O3, ClONO2, HNO3, CH4
  etc) for available data period