Subject: Re: ECCO2 output
Date: Fri, 25 Apr 2008 17:50:13 +0100
From: Alastair Gemmell <alg@mail.nerc-essc.ac.uk>
To: Dimitris Menemenlis <menemenlis@jpl.nasa.gov>

Hi Dimitris,

I have some results from analysing your ECCO2 netcdf output files in OceanDIVA.

In my talk at the Ocean Sciences Meeting in Florida, I presented some
results of comparing ocean syntheses.  In particular, the talk
contained data on September 2004.  I have run OceanDIVA on ECCO2 for
September 2004 also, and I have put images of PDFs in a zipped
directory available from my web space.  In this directory there are
also sub-directories containing equivalent PDFs for ECCO-GODAE,
ECCO-SIO, ECCO-JPL and GECCO, for you to compare with.

If you open one of the .html files within a synthesis sub-directory,
this should show you all the relevant PDF images.  They are sorted by
ocean region, and the observations are shown on the left, with the
misfits on the right.  Both T and S are shown.  If you click on the
'toggle lines' button, then a new image is shown - one which has lines
representing the mean misfit, and +/- one standard deviation
superimposed on the data.

The misfit PDFs are model-obs so the right of the PDF is model too
warm/salty, and the left is model too cold/fresh.

The colours represent data density, so red colours mean lots of data
in a given bin, and blues mean less data.

I have added the ECCO2 data to the online version of OceanDIVA at
http://www.reading.ac.uk/oceandiva
However, although it works fine on the version on my laptop, it seems
that our server is having memory problems with ECCO2 on the online
version of OceanDIVA.  It is the highest resolution model which we
have access to, so this seems to be causing problems.  Our main web
server technical guy is away today, but I will work with him next week
to fix the problem (it works fine on my laptop, so I'm sure we can get
it working on the web server too!)

I'll let you know when its up and running properly.

The URL to download the PDF analyses is:

http://www.nerc-essc.ac.uk/~alg/ECCO2_OceanDIVA_analysis.tgz

and, my talk from the Ocean Sciences meeting is available at the
bottom of the OceanDIVA information page, which is linked to from the
OceanDIVA interface (http://www.reading.ac.uk/oceandiva).

I hope this is of some help to you, and I hope to get ECCO2 working in
the online version of OceanDIVA next week, so that you can generate
any PDFs that you wish to, rather than just the Sept 2004 ones which I
have made at the moment.

Feel free to email me with any questions, if this email isn't very
clear, or any of the links don't work.

Cheers
Alastair