Wrong diagnostic for sea ice extent in common_ocean_post.txt

The diagnostic for sea ice extent ("sie") is not calculated according to the sea ice community's standards. The National Snow and Ice Data Center defines sea ice extent as

SUM( grid_cell_area WHERE sea_ice_concentration > 0.15 ) (http://nsidc.org/cryosphere/seaice/data/terminology.html)

whereas the current diagnostic in ocean_diagnostics is

SUM (grid_cell_area * sea_ice_concentration WHERE sea_ice_concentration > 0.15) (common_ocean_post.txt, line 804: variable ${d}Extend is read from the output of the program compiled from /cfu/software/cdftools2.1/cdficediags.f90; both 1) and 2) diagnostics are calculated, but ?Extend corresponds to the wrong one).

To do: ${d}Extend has to be changed into ${d}Exnsidc in common_ocean_post.txt

Impacts

This can make a very large difference: on a random NetCDF I have about 2.3 million km² difference in the Southern Hemisphere, 0.6 million km² in the Northern Hemisphere. This is actually critical in summer.

Northern Hemisphere NVolume (10^9 m3) 29254.219726848001
NArea (10^9 m2) 12637.677027328000
NExtend (10^9 m2) 12603.235500032000
NExnsidc (10^9 m2) 13283.136372736000

Southern Hemisphere SVolume (10^9 m3) 3912.7739269120002
SArea (10^9 m2) 10679.136288768001
SExtend (10^9 m2) 10458.337640448000
SExnsidc (10^9 m2) 12769.259683840000