Hi all,

I would like chlor_a mapped to a 360(rows) x 720(columns) grid. The 'l3m_data' mapped data files are already gridded to a 2160 x 4320 grid. According to (albeit older) documentation, this is 1 x 1 degree global grid, how can this be so?

So my questions are:

What level data would be best for this? Essentially I just need chlor_a data and some latitude/longitude variables/information.
I can see the level-2 data has this info, but I can't quite see just yet how binning into a long vector (4000 odd entries I think..) corresponds to a gridded dataset?
Can SeaDAS regrid to a different 'resolution' grid?

Any help, pointers to documentation would be great.

Regards,

Dom.

Hi Dom,

The l3m_data you refer to represent a global grid in which each pixel
covers an area that is 5 minutes of longitude by 5 minutes of latitude.
I'm not sure which documentation you refer to.

If you want a 1-degree grid you can just average pixels in 12-by-12 groups
in the 5-minute data (which we nominally refer to as 9-kilometer data).

Please see the following for descriptions of our various archive products.
http://oceancolor.gsfc.nasa.gov/DOCS/ocformats.html

Regards,
Norman

Norman,

Thanks, that clears things up, and the averaging shouldn't be too hard to do.

I was sure I was missing some fields when I read in the data - but it literally is just the array of chlor_a right?

I'll try and find the document, it's an old algorithm paper from around 2005 I think. Perhaps if I find it, we might resolve my misunderstanding further.

Regards, and thanks,

Dom.

Dom,

Yes, the l3m_data segments of data files with names like
A20100322010059.L3m_MO_CHL_chlor_a_9km contain a
single 4320 by 2160 array of 32-bit floating point values
that represent chlorophyll-a concentrations in milligrams
per cubic meter.    You can find the "no data" value (currently
-32767.0) listed along with other useful metadata in the
numerous other attributes stored in the file.

Norman

Hi Norman,

Ok - I seem to have had luck regridding (after remembering not to average NaN values!).

The paper I mentioned is J. Acker et al., 2002 ('SeaWiFS Ocean Color Data Arxchive and Distribution System..')- the table on page 2 of the paper states that the Level 3 SMI has a 1x1 deg. file format... Am I missing something here, or is this just outdated now? I imagine the data resolution has changed over time..?

Regards,

Dom.

Hi Dom,

Earlier in the SeaWiFS mission we were producing SMI products that
were 4096 by 2048 pixels.  We switched to 4320 by 2160 to address
a spatial aliasing issue that we were seeing in the data.  I do not
remember that we ever produced a one-degree SeaWiFS archive
product, but then I have forgotten other things over the years. 

It may be that Table 1 (page 91) of the paper you mention refers to
some of the products distributed by Giovanni which is supported
by a separate group from ours -- the first author being a primary
developer of that interface.

At any rate, you should be referring to the format descriptions
I alluded to in my first response if you are downloading products
from our (NASA's Ocean Biology Processing Group) archive.

Regards,
Norman