Back
Readme
1. Introduction
This file provides information to OCMIP participants describing
how to
-
retrieve data files as necessary to force carbon-cycle simulations
that are to be run according to standard OCMIP protocol (as initially discussed
at the first OCMIP meeting 20 January 1995, Hamburg, Germany), and
-
send model results to LSCE where the standard analysis of
OCMIP will take place.
1.1 File Transfer
All files mentioned below are available by ftp from LSCE.
To transfer these files used for the standard OCMIP simulations, do the
following:
-
issue the ftp command,
-
type the appropriate name and password (as prompted), and
-
continue by typing only the underlined portions below, with
a RETURN after each.
$ ftp asterix.saclay.cea.fr
Name: ocmip
Password: xxxxxxxx
ftp> cd /home/geo/ocmip
ftp> prompt
ftp> binary
ftp> mget guide* README* *.f *.dat *.Z
ftp> quit
After transfer is complete, uncompress any files ending
with ".Z" with
uncompress *.Z
The resulting invoice of files follows:
1.2 Timetable
Participants are asked to send model results for Phase A
(alias Anything goes) as quickly as possible. We desperately need your
results by the end of May, if the presentation at the First GAIM Science
Conference in September is to be a success. In the spirit of Phase A, we
ask that if you already have model results, please send them immediately.
As for Phase B (where simulations must be forced with the standard model
input specified below), that part of OCMIP does not officially start until
just after the GAIM conference. Nonetheless, please begin Phase B as soon
as possible and send your results when they become available. This will
help avoid duplication of analytical efforts.
2. Model input
2.1 Subroutine (Fortran)
-
funcchem.f
Fortran functions to calculate equilibrium constants
and the solubility of carbon dioxide in sea water. Taken from the U.S.
Department of Energy's "Handbook of Methods for the Analysis of the Various
Parameters of the Carbon Dioxide System in Sea Water", version 2, ed. by
A.G. Dickson and C. Goyet, September, 1994. (Prepared for the U.S. Dept.
of Energy, Special Research Grant Program 89-7A: Global Survey of Carbon
dioxide in the Oceans). These functions should be used to describe carbonate
chemistry for OCMIP simulations of (1) the solubility pump and (2) the
solubility + biological pumps.
2.2 Atmospheric Records
-
c14sth.dat,
c14equ.dat, c14nth.dat
Atmospheric C-14 levels for three latitudinal bands:
southern hemisphere (90S-20S), equatorial (20S-20N) and northern hemisphere
(20N-90N). These records are the same as those supplied by M. Heimann for
the IPCC-directed evaluation of future emissions and concentrations of
carbon dioxide (Enting et al., 1994)
-
splco2.dat
From Enting et al. (1994): "Spline fit to Mauna Loa plus
ice core data. Standard case, growth rate of 1.7 ppmv/yr for 1990.5. Columns
are time, CO2, d/dt of CO2. Time intervals 0.5 years 1765.0 to 1990.5"
-
stab.dat
From Enting et al. (1994): "Prescribed CO2 concentrations
for the 7 stabilisation scenarios. Half-year intervals."
2.3 Surface Fields
-
Kgw.cssmi.lev.ieee.Z
Gas transfer coefficient (i.e., solubility*piston velocity)
in [mol/(m^2 yr uatm)] on the Levitus grid. The piston velocity K is computed
from an adapted form of Wanninkhof's (1992) equation (8), taking into account
additional information available from the the satellite SSM/I wind measurements.
With Wanninkhof's equation (8) and equation (3) from Boutin and Etcheto
(see README.satdat), we can use the satellite derived monthly average of
the variance <v> and the average of the square of the monthly wind speed
<<u>2> so that
K = (b(t) + d * ( <v>+ <<u>2>)) *
(Sc(t)/660)**(-0.5)
where b(t) is Wanninkhof's term for the chemical enhancement
and Sc(t) is the Schmidt number. Both values depend upon temperature, for
which we have used monthly surface maps from Levitus (1982).
The coefficient d was adjusted to obtain an average value
of the CO2 gas transfer coefficient Kg (=K*S) of 0.061 mol/(m^2 yr uatm)
as obtained by Broecker et al. (1986). For the CO2 solubility S, we employ
the equation from Weiss (1974) (see funcchem.f, function "CO2smoll") and
again, surface temperatures from Levitus (1982).
The average value of Kg is computed only over the ice-free
ocean as determined using the seaice cover climatology consolidated at
GFDL (B. Samuels, pers. comm.) from data of Walsh (1978) and Zwally et
al. (1983). This analysis yields d = 0.315 .
We wish to avoid problems associated with grid transformation
or dataset inconsistencies (i.e., absence of wind data over ocean grid
points not completely covered by sea ice, possible because wind speed and
sea ice data do not derive from the same source). Thus, we first made all
computations on the original 2.5 x 2.5 grid of Boutin and Etcheto. Next
we filled missing values, (including both land and ocean grid boxes having
100% ice cover according to the satellite SSM/I climatology--see README.satdat),
with the average of adjacent ocean values (i.e., ocean grid boxes without
100% sea ice cover). Finally we interpolated to the Levitus 1 x 1 grid
(using an algorithm which considers the relative area that each source
grid box contributes to each target grid box, which in this case is superior
to linear interpolation).
-
fice.cwz.lev.ieee
Fractional sea ice cover (on the Levitus grid) consolidated
from monthly sea ice concentrations in the N. hemisphere (Walsh, 1978)
and in the S. hemisphere (Zwally et al., 1983). This dataset was kindly
made available through B. Samuels (GFDL), who describes the consolidation
as being performed "using techniques developed by Broccoli (GFDL, personal
communication). Sea ice concentrations range from 0.0 to 1.0, where concentrations
greater than or equal to 0.2 are considered to be sea ice and concentrations
less than 0.2 are open water. Land values have been filled by a linear
extrapolation technique to eliminate problems associated with grid transformation."
To be consistent, this dataset is given here for OCMIP on the Levitus grid.
Otherwise, it is unchanged from the original dataset of B. Samuels (which
is also 1 x 1, but with longitudes spanning 0-360 degrees).
=> SPECIAL NOTE: for OCMIP, each participant should reset
all values below 0.2 to zero before proceeding with model simulations.
-
land.mask.lev.ieee
The land mask used by Levitus (1982): for land, mask(i,j)=0.0;
for ocean, mask(i,j)=1.0 .
2.4 File Format
All files above were written as IEEE 32-bit binary; they
are given on the standard "Levitus" grid (1 x 1 degree, having the first
grid box (1,1) centered at -179.5 and -89.5 (i.e., 179.5W, 89.5S) and the
last grid box (360,180) centered at +179.5, +89.5). Below is an example
of the Fortran necessary to read OCMIP surface fields.
character*80 filename filename='Kgw.cssmi.lev.ieee'
open(unit= 60, file=filename, form='unformatted')
c For Cray only, uncomment following line to signal IEEE
binary
c call asnfile(filename,'-N ieee -F f77',ier)
c Read data from January (mo=1) to December (mo=12)
do mo = 1,12
read(60) ((Kgw(i,j,mo), i=1,360), j=1,180)
end do
2.5 Plots of surface fields
PostScript plots of surface fields used for OCMIP can be
found in the plots subdirectory. To
visualize results, you'll need to first transfer them to your machine (by
ftp as described in Section 1.1).
Each plot's filename has 5 fields: the first 3 fields
describe the file contents as defined below; the last two fields do not
change.
AAA.BBB.CCC.ps.Z
AAA = Data Type
BBB = Data Source
CCC = Grid
ps = PostScript
Z = Compressed file
Data Type:
-
fice = Fractional sea ice cover (values between 0.0 and 1.0)
-
pvw = Piston velocity [cm/hr] from Wanninkhof's equation
(8), satellite winds (See section 2.5), and Levitus temp.
-
Kgw = Gas transfer Coefficient [mol/ (m^2 yr uatm)], the
product of K*S, where K is from the Wanninkhof piston velocity (shown in
previous map) and S is the CO2 solubility from Weiss (1974) using Levitus
surface temperatures
-
wind = Wind Speed [m/s]
Note: at the end of each "Data Type" specification, there
are two additional pieces of information:
-
m signifies that results are masked with both land and ice
information (here land and ice mask are colored differently); no "m" means
all land grid boxes and all ocean grid boxes having 100% ice cover have
been filled with the average of adjacent ocean grid boxes that do not have
100% ice cover (as is the case for binary data files given on the Levitus
grid).
-
01 signifies results plotted for the 1st quarter of the year
(i.e., January, February, and March);
-
likewise 02, 03, and 04 signify quarters 2, 3, and 4, respectively.
Data Source:
-
cwz = Climatology from the consolidated data of Walsh (1978)
and Zwalley et al. (1983)
-
cssmi = Satellite (SSM/I) derived fields from Boutin and
Etcheto (see section 2.3 and README.satdat)
Grid:
-
lev = Levitus 1 x 1 degree grid
-
2_5 = Original grid for Satellite derived data
3. Model output
This section is designed to answer the following question:
How do I organize my model results so that they can be
analyzed at LSCE?
To facilitate the transfer of model results to LSCE for
analysis, we avoid asking participants to produce data files according
to a rigid structure. Instead, it appears much simpler if each participant
just copies two subroutines (gridlab.f and trunlab.f), making only a minimum
of changes (read statements, etc.) in order to pass his results as arguments
to these two "standard" OCMIP routines.
A detailed explanation of the necessary changes are given
as comments in each subroutine. Your modified versions of these subroutines
will be employed at LSCE along with the files containing your model output,
in order to transfer your results, iin a standard fashion, to the programs
used for OCMIP analysis. After your model output is ready and you have
modified the two subroutines mentioned above, please let me know
(orr@cea.fr) ASAP. Then I will create
a special directory into which you may transfer
your results by ftp. Alternatively, we can transfer your results to LSCE
(i.e., if you make them available by anonymous ftp).
As described in the OCMIP guidelines, all participants
are requested to send their results on their original grid. For results
from models that do NOT use a rectangular grid (i.e., only Hamburg and
LODyC models, for the moment), we ask that you also duplicate all your
results on a 2 x 2 degree rectangular grid (spanning from -180 to 180 Longitude,
-90 to 90 Latitude); along with this additional output, please send modified
versions of the data transfer subroutines We regret the extra effort this
may require and hope that in the near future this will not be necessary.
References
Enting, I.G., T. M. L. Wigley, M. Heimann, 1994.
Future Emissions and concentrations of carbon dioxide:
key ocean / atmosphere / land analyses, CSIRO Aust. Div. Atmos. Res. Tech.
Pap. No. 31, 118 pp.
Levitus, S., 1982.
Climatological atlas of the World Ocean, NOAA Prof. Pap.
13, U.S. GPO., Washington, D.C., 173 pp. Walsh, J. 1978. A data set on
northern hemisphere sea ice extent, 1953-1976. Glaciological Data, World
Data Center for Glaciology (Snow and Ice), Report GD-2, 49-51.
Wanninkhof, R., 1992.
Relationship between wind speed and gas exchange over
the ocean, J. Geophys. Res., 97, 7373-7382 Weiss, R. F., 1974. Carbon dioxide
in water and seawater: the solubility of a non-ideal gas, Marine Chem.,
2, 203-215.
Zwally, H. J., J. Comiso, C. Parkinson, W. Campbell, F. Carsey,
and P. Gloerson, 1983. Antarctic Sea Ice, 1973-1976: Satellite Passive Microwave
Observations, NASA, 206 pp.
Contacts
orr@cea.fr
Back