Hello again,

2 things in this message:

1. info about a new web page on RUC2 products
2. changes in RUC2 this morning

----------------
1. There is a new on-line web page for description of how RUC2 diagnostic variables are calculated. This was discussed in the last teleconference on 15 Dec 97 and I think I heard several people say that this would be helpful, and there is something to look at now at

http://maps.fsl.noaa.gov/vartxt.cgi

There are also links to this in the 40km RUC2 description at

http://maps.fsl.noaa.gov/MAPS.40km.html

Not everything is on this page yet, but there is information there on precip type, CAPE/CIN, LI/best LI, snow accumulation, precipitation amount, etc. that some of you may find useful.

Please suggest other things to add (including at the teleconference).

2. There were a couple of small changes made to RUC2 this morning (before the 12Z run).

Precipitation type algorithm fixed - While writing the web page description mentioned above, a bug was found in the code itself. This bug was short-changing the IP (ice pellet) indicator and rain, and was showing up too much snow in the 0-2 deg C range. By the way, this is mentioned in the web page
description of precip type, but be aware that the RUC2 can have mixed precip types; i.e., more than one indicator may be turned on at a given grid point.

Snow cover is now changing - Because the GRIB precision for snow depth had been set previously to 10 cm, the snow cover field had not been changing for a few weeks since the 1-h snow accumulation (for the 1-h cycle) is usually less than 10 cm). So there could have been some low-level temperature problems in areas that have more recently acquired snow cover. As of 12Z this morning, the RUC2 started using a higher precision GRIB value for snow depth, and the snow depth from the 40km MAPS (which was correct since GRIB is not used at FSL for cycling) was copied to NCEP. The RUC2 analysis was forced to use an Eta forecast at 12Z this morning for its guess field as part of correcting the snow depth. Usually, of course, the RUC2 analysis would use its own previous 1-h forecast for its guess field.

Stan