NOAA

Geophysical Fluid
Dynamics Laboratory

Skip to: [content] [navigation]
search gfdl:

If you are using Navigator 4.x or Internet Explorer 4.x, this site will not render correctly!

gfdl's home page > data portal > deccen coupled climate models > CM2.x coupled climate models > CM2.X model ocean and sea ice grids information

CM2.X model ocean and sea ice grids information


[NOAA bullet] CM2.X model ocean and sea ice grids information.

Analyzing model output on the interpolated grid

Because users often find analyzing GFDL's CM2.X ocean and sea ice model output quite challenging when the data is stored on the models' native tripolar grid, we have begun to make some (but not all) of these variables available for download from files that have had the variables interpolated to a rectangular grid. The first interpolated files to be made available for download are scalar 2D variables. Next will be scalar 3D variables. Later, we hope to make interpolated 2D and 3D vector quantities available.

Considering 2 dimensional scalars

[seaarea fraction fig.]

The two-dimensional GFDL CM2.1 ocean and sea ice fields have been interpolated conservatively to a longitude-latitude grid north of 65N. South of this latitude, the computational grid is longitude-latitude and so no interpolation was necessary. The grid cells of the ocean/ice computational grid are, of course, either entirely ocean or entirely land. After interpolation, however, some of the longitude-latitude grid cells north of 65N are partially ocean covered and partially land. (This can be seen in the polarstereographic figure of ocean area fraction.) Therefore, in this region, it is necessary to apply ocean-fraction weighting when integrating or averaging. These weights are contained in the variable seaarea.

To integrate a 2D scalar field S accurately, compute:
 
    integral (S * seaarea) dA
 
To average S accurately, compute:
 
    ( 1 / A integral [ S * seaarea ] dA ) division ( 1 / A integral [ seaarea ] dA )
 
Failure to weight with seaarea when averaging or integrating north of 65N has the effect of giving excessive weight to coastal cells with partial land (see figure). This can lead to significant errors north of 65N.


[error graph] For example, consider the impact of calculating northern hemisphere sea ice area with and without weighting. There is a seasonally dependent high bias in the unweighted field due to overvaluing the ice cover adjacent to land areas north of 65N, as shown in this graph.

(NOTE: South of 65N the ocean and sea ice models' native grid follows a rectangular latitude by longitude arrangement that can be easily ingested by most analysis software. Thus, the grids and data values found in the tripolar and interpolated grid files are identical south of 65N. CM2.X ocean and sea ice models output on the tripolar grid can only be found on this GFDL Data Portal, and not at PCMDI's IPCC WG1 data archive)

Questions related to the GFDL CM2.x models may be directed to…


[email GFDL.Climate.Model.Info at noaa dot gov]