Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

trans_ch4ctl_acpd_suppl

VIEWS: 5 PAGES: 33

									Supplementary information of “TransCom model simulations of
CH4 and related species: Linking transport, surface flux and
chemical loss with CH4 variability in troposphere and lower
stratosphere”

Table S1: List of responsible scientists and e-mail addresses for different chemistry-
transport model simulations.
 Model                     Institution            Modeler Name
 Name
 ACCESS          CSIRO Marine and                 K. D. Corbin
                 Atmospheric Research,            <kdcorbin@atmos.colostate.edu >,
                 Australia                        R. M. Law <Rachel.Law@csiro.au>
 ACTM            Research Institute for Global    R. Saito <rsaito@jamstec.go.jp>,
                 Change, Japan                    A. Ito <itoh@nies.go.jp>,
                                                  P. K. Patra<prabir@jamstec.go.jp>
 CAM-Chem        Cornell University, USA          L. Meng <meng2000@gmail.com>,
                                                  P. G. Hess <pgh25@cornell.edu>
 CCAM            CSIRO Marine and                 Z. Loh <Zoe.Loh@csiro.au>,
                 Atmospheric Research,            R. M. Law <Rachel.Law@csiro.au>
                 Australia
 GEOS-Chem       University of Edinburg, UK       A. Fraser <ac.fraser@ed.ac.uk>,
                                                  P. I. Palmer <pip@ed.ac.uk>
 IMPACT          Lawrence Livermore               D. Bergmann <Bergmann1@llnl.gov>,
                 National Laboratory, USA         P. Cameron-Smith <pjc@llnl.gov>
 LMDZ            Institut Pierre Simon Laplace    A. Fortems-Cheiney
                 des sciences de                  <audrey.fortems@lsce.ipsl.fr>,
                 l'environnement, France          P. Bousquet <bousquet@lsce.ipsl.fr>
 MOZART          Massachusetts Institute of       M. Rigby <mrigby@mit.edu>,
                 Technology, USA                  R. G. Prinn <rprinn@mit.edu>
 NIES08i         National Institute for           D. Belikov
                 Environmental Studies, Japan     <dmitry.belikov@nies.go.jp>,
                                                  S. Maksyutov <shamil@nies.go.jp>,
 PCTM            NASA Goddard Space Flight        H. Bian <huisheng.bian-1@nasa.gov>
                 Center, USA                      S. R. Kawa
                                                  <stephan.r.kawa@nasa.gov>
 TM5             SRON Netherlands Institute       S. Houweling <s.houweling@uu.nl>,
                 for Space Research, The          M. Krol <M.C.Krol@uu.nl>
                 Netherlands
 TOMCAT          University of Leeds              C. Wilson <c.wilson@see.leeds.ac.uk>
                                                  E. Gloor <E.Gloor@leeds.ac.uk>
                                                  M. P. Chipperfield
                                                  <martyn@env.leeds.ac.uk>
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                    2


Table S2: Details of data sources and responsible organizations for taking
measurements of CH4, MCF and SF6 at 8 different baseline monitoring stations under
the AGAGE [Cunnold et al., 2002; Prinn et al., 2005] and NOAA [Dlugokencky et al.,
1998; Butler et al., 2004] networks.


Station name & location                       Data network & managing institution
ALT, Alert, Canada;                           NOAA: Global Monitoring Division, ESRL
  o       o
62 W, 82 N, 210m                              (Edward Dlugokencky; James Elkins)
BRW, Point Barrow, USA;                       NOAA: Global Monitoring Division, ESRL
157oW, 71oN, 11m                              (Edward Dlugokencky; James Elkins)
MHD, Mace Head, Ireland;                      AGAGE: University of Bristol
10oW, 53oN, 25m                               (Simon O’Doherty; Peter Simmonds)
MLO, Mauna Loa, Hawaii, USA;                  NOAA: Global Monitoring Division, ESRL
      o       o
156 W, 20 N, 3397m                            (Edward Dlugokencky; James Elkins)
RPB, Ragged Point, Barbados;                  AGAGE: University of California, San
59oW, 13oN, 45m                               Diego (Ray Weiss)
SMO, Samoa, USA;                              AGAGE: University of California, San
      o       o
171 W, 14 S, 42m                              Diego (Ray Weiss)
CGO, Cape Grim, Australia;                    AGAGE: Commonwealth Scientific and
      o   o
145 E, 41 S, 94m                              Industrial Research Organization
                                              (Paul Fraser, Paul Steele; Paul Krummel)
SPO, South Pole, Antarctica;                  NOAA: Global Monitoring Division, ESRL
  o       o
25 W, 90 S, 2810m                             (Edward Dlugokencky; James Elkins)




DRAFT                                  30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                  3




Figure S1: Latitude-pressure of ACE-FTS, HALOE/UARS and TransCom simulated
CH4 in the upper troposphere and lower stratosphere. This plot is similar to Fig. 3, but
focus is given for the stratospheric altitudes and HALOE observation is include while
ACTM_OH (similar distribution as the ACTM) is not shown.


DRAFT                              30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison             4




Figure S2: Latitude-pressure 222Rn distribution along the 70oE longitude (South
Asian monsoon domain) for the averages during Dec-Jan-Feb (DJF) months of the
year 2003-2004. Note the unequal colour bar. LMDZ model did not submit 222Rn
simulation results.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                 5




Figure S3: Latitude-pressure 222Rn distribution along the 180oE longitude (over the
central Pacific Ocean) for the averages during DJF months of the year 2003-2004.
Note the unequal colour bar.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison              6




Figure S4: Latitude-pressure 222Rn distribution along the 70oE longitude for the
averages during Jun-Jul-Aug (JJA) months of the year 2003. Note the unequal colour
bar.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison               7




Figure S5: Latitude-pressure 222Rn distribution along the 180oE longitude for the
averages during JJA months of the year 2003. Note the unequal colour bar.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison            8




Figure S6: Latitude-pressure SF6 distribution along the 70oE longitude for the
averages during DJF months of the year 2003-2004.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison             9




Figure S7: Latitude-pressure SF6 distribution along the 180oE longitude for the
averages during DJF months of the year 2003-2004.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison            10




Figure S8: Latitude-pressure SF6 distribution along the 70oE longitude for the
averages during JJA months of the year 2003.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison             11




Figure S9: Latitude-pressure SF6 distribution along the 180oE longitude for the
averages during JJA months of the year 2003.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison             12




Figure S10: Latitude-pressure CH4 distribution along the 70oE longitude for the
averages during DJF months of the year 2003-2004. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison              13




Figure S11: Latitude-pressure CH4 distribution along the 180oE longitude for the
averages during DJF months of the year 2003-2004. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison             14




Figure S12: Latitude-pressure CH4 distribution along the 70oE longitude for the
averages during JJA months of the year 2003. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison              15




Figure S13: Latitude-pressure CH4 distribution along the 180oE longitude for the
averages during JJA months of the year 2003. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison            16




Figure S14: Latitude-pressure CH3CCl3 distribution along the 180oE longitude for the
averages during JJA months of the year 1993. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison            17




Figure S15: Latitude-pressure CH3CCl3 distribution along the 180oE longitude for the
averages during DJF months of the year 1993-1994. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison            18




Figure S16: Latitude-pressure CH3CCl3 distribution along the 180oE longitude for the
averages during JJA months of the year 2003. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison            19




Figure S17: Latitude-pressure CH3CCl3 distribution along the 180oE longitude for the
averages during DJF months of the year 2003-2004. Note the unequal colour bar.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                    20




Figure S18: Annual mean timeseries at MLO and CGO suggesting that all models
started with relatively similar initial conditions for CH4, but drifted away with time
depending on the model behaviour. Fig. S19 shows the simulated time series
integrated for the whole troposphere at monthly time intervals.




DRAFT                              30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison              21




Figure S19: Monthly-mean model values integrated for the troposphere (850-200 mb)
for CH4_CLT (top) and CH3CCl3 (bottom). This is also to show that the initial values
of model played relatively minor role for the CH4 and CH3CCl3 model-to-model
differences. Note the model spread increased significantly from 1990 to 1993 for
CH3CCl3 or to the end of the simulation for CH4.


DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison       22




Figure S20: Same as Figure 6, but for CH4_CTL_E4 (left column) and CH4_BB
(right column) tracers.




DRAFT                          30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison       23




Figure S21: Same as Figure 6, but for CH4_WL_BB (left column) and CH4_INV
(right column) tracers.




DRAFT                          30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                24




Figure S22: Comparison of CH3CCl3 lifetimes calculated using ACTM
photochemical loss rate at each model grid (black line; symbols) and Eqn. 5 of the
main text for ACTMs (green & blue lines). The median and range of all models are
also shown for a reference (red lines). The average lifetimes over the 2000-2007
period using Eq. 5 and aggregating grided ACTM loss rates are 4.60 and 4.59,
respectively.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison             25




Figure S23: Same as Fig. 9, but the results using 3 other CH4 emissions are shown for
the 1998-2002 period. CH4_CTL_E4 is excluded because of its similarity with
CH4_CTL for IH gradients simulation.




DRAFT                            30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                26




Figure S24. Relationships between average CH4 and CH3CCl3 growth rates in
comparison with observations over 8 sites over the 5-years averaging period of 1995-
1999 and 2003-2007. The CCAM, LMDZ and MOZART models are excluded from
the linear fittings in the bottom panel. These three models formulate a different group
with low CH3CCl3 growth rate, probably due to enhanced photodissociation because
LMDZ and MOZART have relatively high concentrations in the stratospheric
altitudes (reason unknown for CCAM).




DRAFT                              30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                 27




Figure S25. Meridional gradients of CH4 in the lower stratosphere (70-30mb height),
(top panel) between Eq-10oN and 20-30oN, and (bottom panel) between Eq-20oN and
30-50oN for models and HALOE. These gradients are fairly symmetrical in both the
hemispheres, and are an indicator of leakiness in the ‘pipe’ model of stratospheric
transport.




DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison              28


We briefly describe the chemistry-transport models used in this
study (incomplete right now due to lack of information – will be
deleted if not completed by the submission time).


http://www.esrl.noaa.gov/gmd/ccl/scales/CH3CCl3_scale.html (08 June 2011)


New CH3CCl3 standards were prepared in 2001-2003 from new reagent material
(99.8% purity). The 1996 scale was based on standards made from reagent now
known to contain several impurities (approx 94% pure). The interim 2002 scale was
based on these new standards, but was not well-defined at mixing ratios less than 40
ppt. The following table defines the 2003 scale. Two equations are used to convert the
1996 scale and the interim 2002 scale to the current 2003 scale.


CH3CCl3 calibration standards, updated: October 2003
                      year           prepared       assigned       residual
                                     (ppt)          (ppt)          (ppt)
ALM-64608B            2001           99.1           99.5           0.4
ALM-65990A            2002           124.8          125.1          0.3
ALM-33800A            2002           104.0          102.7          -1.4
ALM-38422A            2002           187.2          187.1          -0.1
SX-3506               2002           76.7           76.6           -0.1
SX-3518               2003           20.7           20.6           0.0
SX-3519               2003           40.8           40.5           -0.2
SX-3522               2003           33.8           33.9           0.1
std dev                                                            0.58


Convert 2002 scale to 2003 scale
Y = -2.0022e-4*X^2 + 1.0402*X - 2.68
where X = conc (ppt) from 2002 scale


convert 1996 scale to 2003 scale
Y = -1.6120e-4*X^2 + 0.9654*X + 0.2
where X = conc (ppt) from 1996 scale



DRAFT                              30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                 29



SF6
Update: April, 2009
The NOAA-2006 SF6 scale was recently developed and supercedes the 2000 scale.
Several standards that defined the 2000 scale were nearing the end of their useful lives
and needed to be replaced.
The 2006 scale is based on 16 primary standards prepared by static gravimetric
dilution from 99.99% pure SF6 in 5.9-L aluminum cylinders. All standards contain
CO2 and N2O at near ambient levels. Assigned values were determined by fitting a
line to the response of each standard (relative to a 5.45 ppt natural air secondary
standard) measured using an HP6890 GC with electron capture detection. Analysis
precision was typically less than 1% (one standard deviation).
The 2006 scale is 0.033 ppt higher than the 2000 scale at 4 ppt.
To update the 2000 scale to the 2006 scale use:

Y = 4.8546e-3 * X^2 + 9.3479e-1 * X + 0.21664 (where Y = 2006 scale, X = 2000

scale).
2000 scale       2006 scale         difference
                1             1.156                0.156
                2             2.106                0.106
                3             3.065                0.065
                4             4.033                0.033
                5             5.012                0.012
                6             6.000                0.000


The main advantage of these new standards over the previous set is that they have
enabled us to confirm that the ECD used for SF6 analysis has a linear reponse (0-10
ppt) and a zero intercept. A zero intercept was not evident with the previous set of
standards.
The reproducibility of SF6 calibrations was evaluated by measuring five secondary
standards (ranging ftom 4.5 to 7.8 ppt) every 2-3 weeks. Treating these standards as
unknowns, the assigned mixing ratios have varied less than 0.04 ppt over the last 2
years. Therefore, we estimate the reproducibility of SF6 analysis to be 0.04 ppt.

Standards that define the NOAA-2006 SF6 scale
    Cylinder               Year                  Prepared            Assigned              Residual
                                                  (ppt)                (ppt)                (ppt)
FA-1861                    2000                    2.41                2.39                 -0.02
FA-1878                    2000                    2.91                2.89                 -0.02
FA-1843                    2000                    4.73                4.76                  0.03


DRAFT                              30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                30


FA-1850                    2000                 4.51                  4.53                0.02
FA-1851                    2000                 4.32                  4.35                0.03
FA-1856                    2000                 1.14                  1.15                0.01
FA-1865                    2000                 4.12                  4.13                0.01
FA-2205                    2003                 4.94                  4.92                -0.02
FA-2207                    2003                 5.98                  5.98                0.00
FA-2208                    2003                 6.97                  7.01                0.03
FA-1940                    2005                 1.52                  1.49                -0.03
FA-2139                    2005                 3.13                  3.14                0.01
FA-2557                    2005                 3.86                  3.85                -0.01
FA-2567                    2005                 5.99                  6.02                0.03
FA-2569                    2005                 7.92                  7.90                -0.02
FA-2585                    2005                 9.83                  9.79                -0.04



ACCESS (CSIRO) :
ACCESS is run here as a climate model using observed monthly-mean sea surface
temperatures (SST). It is not nudged to analysed fields, and hence does not have
accurate synoptic variability compared to observed weather patterns. Semi-
Lagrangian advection with an ECMWF monotone quasi-cubic interpolation scheme
and a higher-order scheme using quasi-cubic interpolation in the horizontal and
quintic in the vertical. Adaptive detrainment, deep and mid convection (based on
Gregory and Rowntree, 1990). Turbulent transport for boundary layer as described by
Lock et al. (2000).


ACTM (JAMSTEC) :
Advection scheme: Lin and Rood (1996)
Convection scheme: simplified Arakawa and Schubert (1974)
Boundary layer scheme: Non local scheme
Vertical diffusion: Mellor and Yamada (1982)


ACTMoh (JAMSTEC) : ACTM, but using its own OH concentration, and excluding
SOIL sink. The OH field used in this model simulation is from ACTM's own. The OH
distribution is available in the Supplementary directory on the FTP site. The SOIL


DRAFT                             30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                 31


sink is excluded, because the scaled OH is able to reproduce MCF growth rate, but
CH4 growth rate is underestimated. Please refer to Patra et al. (JMSJ, 2009) for
further details.


CCAM (CSIRO) :


GEOS-Chem (University of Edinburg) :
Model is run for the periods 1990-2006: 30 layers (GEOS4 meteorology), 2007: 47
layers (GEOS5 meteorology) at hybrid sigma-pressure. Approximate height of levels
above the surface: up to 0.002 hPa. Model time step is 15 minutes for transport and
emissions, 60 minutes for chemistry.
Convection scheme:
Boundary layer scheme:
Vertical diffusion:
Horizontal diffusion:


GEOS-Chem_DOH (University of Edinburg) :
Model is run for the periods 1990-2006: 30 layers (GEOS4 meteorology), 2007: 47
layers (GEOS5 meteorology) at hybrid sigma-pressure. Approximate height of levels
above the surface: up to 0.002 hPa. Model time step is 15 minutes for transport and
emissions, 60 minutes for chemistry.
Convection scheme:
Boundary layer scheme:
Vertical diffusion:
Horizontal diffusion:


IMPACT (LLNL) : IMPACT is developed at Lawrence Livermore National
Laboratory and run at horizontal resolution: 46 latitudes by 72 longitudes (4x5
degree), and vertical resolution (number of levels and type e.g. sigma, hybrid): 55
hybrid. Approximate height of levels above the surface: 65 m (center of lowest grid
box). Model time step: minimum 60 min. Meteorology is driven by GEOS4-ceres
reanalysis - 1x1.25 degree analysis for U, V, T, PBL, clouds, Psf, kzz, convective
mass flux 6 hourly time intervals.
Advection scheme: Lin and Rood (1996)


DRAFT                                30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison                  32


Convection scheme: Rasch et al. JGR 1997
Boundary layer scheme: Apply large values of kzz within pbl (from met data)
Vertical diffusion: Walton et al. JGR 1988
Reference for model: Rotman et al., JGR, 2004


IMPACT_1x1.25 (LLNL):
Same as IMPACT, but run at higher horizontal resolution of 1 x 1.25 degrees latitude
x longitude grid.


LMDZ (LSCE) :
MOZART (MIT) :
The model is run in offline mode using NCEP/NCAR reanalysis fields at T63 spectral
truncations and 28 vertical levels at 6-hour time intervals. Advection scheme is Lin
and Rood (1996), convection schemes are shallow and mid-level [Hack, 1994], and
deep convection [Zhang and MacFarlane, 1995]. Boundary layer scheme is from
Holstlag and Boville [1993. The model is described in Emmons et al. [2010].


NIES08i (NIES):
Offline model uses 6 hourly JCDAS reanalysis winds [Onogi et al., 2007] on 40
model levels interpolated to hybrid sigma-isentropic coordinates. Vertical transport on
14 isentropic layers (above 350 K) is driven by climatological radiative heating-
cooling rate provided by JCDAS and adjusted to match the mean age of the air in the
lower part of the stratosphere. Other features: second-order van Leer advection, deep
convection according to Tiedtke [1989] adjusted to JCDAS rain rate, bulk boundary
layer with 3-hourly PBL height from ERA-Interim reanalysis. Model details except
recently added isentropic coordinates option are described in Belikov et al. [2011].


PCTM (GSFC) :




TM5 (SRON) :




TM5_1x1 (SRON) :


DRAFT                              30 September 2012
Patra et al.: TransCom CH4, SF6, CH3CCl3, 222Rn model intercomparison           33




TOMCAT (University of Leeds) :
Advection scheme: Prather (1986), Convection scheme: Tiedtke (1989), Boundary
layer scheme: Holtslag & Boville (1993)




DRAFT                            30 September 2012

								
To top