Description of the initial conditions and forcing data used for the SCM
SHEBA runs. In most cases the data directories contain meta data files

ITD initial conditions
Data directory: /data/sheba/ghdata
Source: JOSS, PI:  Rothrock

The ITD initial conditions are derived from the SCICEX cruise data.
The data are provided as the average distribution for each of 34
tracks taken over 4 days from Sept 28 - Oct 1.  I computed a single
distribution from these by weighting each track's distribution by the
number of scans taken on that track. The models begins on Oct 22,
1997, which is 22 days after the last SCICEX track. I assume no growth
occured between the SCICEX cruise and Oct 22.

The model calls a routine in ice_init.F that sets these values
directly.

--------------------------------------

Deformation forcing -- Opening and closing rates
Data directory: /data/sheba/open_clos and /data/sheba/lindsay
Source: Harry Stern and Ron Lindsay

Harry gave me opening and closing rates on 28 Sept 2000.  Harry says
he generated the data for "4 sizes of square regions centered on the
SHEBA site.  Within each region there are a certain number of 5x5 km
cells.  The crude method I used is simply to look at the area changes
of the cells within the region.  For example, in the 50x50 km region
there are normally 100 cells.  Each cell has an initial area of 25
km^2, so the total area is 2500 km^2.  Suppose 20 of the cells
increase in area over the time interval between the SAR image
observations, and suppose the total area increase of those 20 cells is
150 km^2.  Then the opening is 150/2500 = 0.06 = 6% per time interval.
If the images are 3 days apart, it's 2%/day."

Also from Harry, "The method probably overestimates opening and
closing because small errors in tracking the ice motion cause some
cells to "open" and others to "close" erroneously.  Also, because of
the square geometry, a shear line running at an angle across the scene
will cause some cells along the line to open and others to close, even
if the actual motion across the shear line is pure sliding."

I have two types of data files that can be used to force the column
model.  The first uses Harry's opening and closing data only. (Does
not assume elliptical yield curve.) I had to fill in two parts of the
timeseries to make it complete. (1) For Sep 15 to Nov 2, 1997, I used
the mean for all data available for period between 15 Sep to Nov 15
from 1997 and 1998. (2) For Dec 23, 1997, to Jan 10, 1998, I used the
mean for all data available for period between Dec 11 to Jan 25.
Finally the data were linearly interpolated from the SAR data
(considered as the value at the midpoint of the time between scans) to
hourly intervals.  Then the data were smoothed with a 12-hour running
mean, which hardly changes them.  The opening and closing rates
are otherwise unmodified. 

file: open_clos_oct22.dat 
day since 12/31/96  opening  closing
                     [1/s]    [1/s]

 
The second option is derived from Harry's divergence and shear data,
also estimated from SAR. This time I computed an opening rate by
assuming an elliptical yield curve with e=2. Also I filled in data
during the two "gaps" described in the previous paragraph with data
from Ron Lindsay. Ron says he used SAR and buoy data, so I suppose
during the gaps he just used buoy data. I used linear interpolation
and 12 hour smoothing as above.  My datafiles have four estimates of
closing. First I have the closing from SAR as in the previous
paragraph, except that I limited it to be no greater than the
divergence (which occured because I filled in the closing with the
mean and yet the divergence gaps were filled in with Ron's
values). The next three estimates of the closing were computed from
delta=sqrt(divergence.^2+0.25*shear.^2)
closing=0.5000*(delta-abs(divergence))-min(divergence,0)
closing=0.2500*(delta-abs(divergence))-min(divergence,0) and
closing=0.1125*(delta-abs(divergence))-min(divergence,0) where the
estimates have decreasing contribution from shearing motion.

file: div_clos_oct22.dat
day since 12/31/96  divergence  closing1  closing2  closing3  closing4
                     [1/s]      [1/s]     [1/s]     [1/s]     [1/s]

--------------------------

Zenith Angle
Data directory /data/sheba/zenith
Depends on ships position
from SPO Moritz /data/sheba/ship_position

The ccm ocean albedo parameterization has a strong zenith angle
dependence for the direct beam component of the down solar. I wrote a
utility program to input the ships position and time and call the CCM
routines "shr_orb_params", "shr_orb_decl", and "shr_orb_cosz". I saved
the output and use it as forcing in the ice_only model.

file: zenith_oct22.dat
The data set is in the following format:
day since 12/31/96     cosine zenith angle


--------------------------

Surface flux forcing data from Judy Curry 
flux_force.dat           % starts on oct 10
simip2_forcing_data.txt  % begins on oct 31

Took flux data from Judy Curry's single column model forcing datasets
web site and the snowfall rate from the SIMIP2 web site. I needed
snowfall rates for Oct 22-30 and the last couple of days of the
datafile and so I used the mean from the first 2000 hours starting Oct
31. 

file: currydata_oct22.dat
The data set is in the following format:
 ws10        airT10        qa10         lwrd        swrd    snofall
[m/s]         [K]          [g/kg]      [W/m^2]     [W/m^2]  [mm/day]
