[Narccap-announce] Major Dataset Update

[Seth McGinnis]

Greetings NARCCAP Users!

We have just completed a major update of the NARCCAP data catalog on
ESG.  Some files have been added, some have been updated, and a few have
been removed.  See below for details.


MM5I Datasets Completed
-----------------------
The MM5I simulations have now been published in their entirety.  All
variables in all tables for all simulations are now available.

Some files from the earlier MM5I runs have also been updated.  In
quality-checking the new files, we discovered that, due to differences
in how the sponge zone was removed, there were two groups of files that
were offset from one another.  (Basically, one group had an extra row of
points along the left and bottom edges, and the other had an extra row
of points along the top and right edges.)  To resolve this issue, we
regenerated ALL the MM5I files to make them match up.

The new files are numerically identical to the old files except along
the outer edges.  If you geolocate your data using lat/lon coordinates,
this issue has no effect on your analysis.  It is only potentially an
issue if you are directly comparing data between two different runs
based on the array index.  We apologize for any confusion generated by
this problem.


SND Replaced with SWE
---------------------
There was some confusion with regard to the variable snd, snow depth,
and whether it was supposed to represent the thickness of the snow layer
or the amount of water in the snow.

We checked the published data and found that, with the exception of the
MM5I-ncep run, it was all snow-water-equivalent, but in some cases the
units were incorrect.

We have updated the files to change the variable from snd (snow depth)
to swe (snow water equivalent) and corrected the units.  (Note: the
MM5I-ncep run now has both snd and swe.  All other runs have only swe.)


HRM3 3D Fill-Ins Published
-----------------------
Various and sundry files and variables were missing from the 3D HRM3
datasets.  They have now been published.  (Note that the variables wa,
cli, and clw are still missing or incomplete for some runs.)


WRFG CGCM3-future Table 3
-------------------------
The Table 3 files from this run were missing every fifth year: each
five-year file held only four years' worth of data.  These gaps have now
been filled in.


ECP2 NCEP Tables 1, 2, and 3
----------------------------
These files were regenerated to correct various issued in the metadata
and ancillary data.  In a few cases, there may be very small numeric
differences in some files due to changes in the postprocessing.  The new
files are not numerically identical, but the differences are small and
should not affect the results of any analysis.


TMSL CCSM 2061 Zeroes Filled
----------------------------
The output from CCSM TMSL (aka CAM3 timeslice) had some all-zero
timesteps in some files for the years 2061-2065.  Those have now been
filled in.


Miscellaneous Updated Files
---------------------------
Some other files may have been updated since they were originally
published.  None of these changes affect the actual model output in the
files in any substantial way, they are just updates to metadata and
ancillary data.


Deprecated Datasets Removed
---------------------------
There were a number of simulations that were published and then re-run
after errors were found in the original runs.  These include the
original ECPC-ncep and WRFP-ncep simulations and the RCM3 runs with bad
lake temperatures.  These datasets have been removed from the ESG
catalog. If for some reason you need data from one of the deprecated
runs, you can request it from us directly.


Our apologies for the long delay in publication of new data.

Cheers,

--Seth

----
Seth McGinnis
NARCCAP Data Manager
Associate Scientist
IMAGe/NCAR
----