rho_un output

[Karen Bice <kbice@xxxxxxxx>]

I have a netcdf file that contains temperature and salinity on an XYZ
grid. I am trying to use the rho_un function to calculate density.

When I say 
let density = rho_un(temp, sal, Z[G=temp])
I get density values that can't be right.

For example, for a specific X,Y location, values look like this (the
first column is depth in cm):
                TEMP    SAL   DENS 
1250     /  1:  27.61  36.76  1019.
3750     /  2:  27.62  36.78  1029.
6250     /  3:  27.64  36.80  1039.
11250    /  4:  27.64  36.82  1057.
18750    /  5:  27.58  36.84  1081.
41250    /  6:  24.83  36.10  1139.
68750    /  7:  20.58  35.46  1192.
111250   /  8:  16.30  35.11  1261.
151250   /  9:  14.24  34.97  1323.
188750   / 10:  13.30  34.90  1384.
238750   / 11:  12.60  34.85  1481.
301250   / 12:  12.17  34.83  1648.

But, more reasonable UNESCO values for in situ densities (in kg/m^3) for
these T,S values would be:
 1250     /  1:  1023.93
 3750     /  2:  1024.04
 6250     /  3:  1024.15
 11250    /  4:  1024.38
 18750    /  5:  1024.73
 41250    /  6:  1025.99
 68750    /  7:  1027.92
 111250   /  8:  1030.60
 151250   /  9:  1032.71
 188750   / 10:  1034.48
 238750   / 11:  1036.73
 301250   / 12:  1039.44

I know that rho_un expects Z in decibars (according to the user's
guide), so I figured that it was using the actual depth values. But even
if that were the case, those depths, interpreted as pressures in
decibars shouldn't yield densities like those in my first list above.

Is the output of rho_un actually something other than kg/m^3?
Can anyone tell me what I am doing wrong here and how I can get good
density values out of rho_un?

Thanks very much.

-- 
Karen L. Bice, Assistant Scientist		Mail Stop #23
Dept. of Geology and Geophysics			phone: (508) 289-3320
Woods Hole Oceanographic Institution		FAX:   (508) 457-2187
Woods Hole, MA		USA 02543-1541