Hi HP-
HansPeter Roesli wrote:
Allow me to raise 2 points in this context.
1. I got just a bit confused about your "The sampling is a weighted
average by default, not the nearest neighbor.". Does this mean that the
sampling follows the setting in Preferences (that has nearest neighbour
as default) or is the sampling for the pseudo-soundings a special case?
My bad, it's neither. Looking at the code, it always uses nearest
neighbor. I think it should use the preference which I was assuming to
be weighted average as the default.
2. May I add that sounding works for either dew point *or* relative
humidity, but *not* for specific humidity.
Send me the formula for computing dewpoint from specific humidity and
temperature. ;-)
Don
Don Murray wrote:
Kevin-
Havener Kevin Civ 14WS/WXED wrote:
Yes, I think the ability to specify a single(/series of) point(s) would
be a nice feature to have. I have frequent reason to probe archived GFS
and ECMWF global analysis fields somewhat randomly, comparing derived
soundings from them to real soundings, satellite soundings, and other
data.
I'll spend some time with Tom and Brendon's advice and see if I can't
make it work for the current project.
At this point, the best solution is to subset the grid over the region
you want to display soundings.
The grid sounding probe was designed to be an interactive probe where
you could drag the probe around easily to get soundings at arbitrary
points. To do this, it needs to read in all the levels for t, dewpoint
(or fields to create it), u and v and perform any calculations to
derive a field like dewpoint, then sample the grid quickly to make the
display. The sampling is a weighted average by default, not the
nearest neighbor.
If we took the approach of specific samplings, we would probably not
be able to do a weighted average approach easily. It would require
some changes to the way we handle the interaction with the underlying
data and the netCDF-Java library which does the reading. We can look
into what it would take to do this at some point.
Don
-----Original Message-----
From: Brendon Hoch [mailto:bhoch@xxxxxxxxxxxx] Sent: Monday, November
02, 2009 8:25 AM
To: Tom Whittaker
Cc: Havener Kevin Civ 14WS/WXED; idvusers@xxxxxxxxxxxxxxxx
Subject: Re: [idvusers] Soundings from global grib file
> I cannot answer the question of how to specify the pointer location,
> in the interactive mode -- other than dragging it to a point. You
> might look at displaying the special "Location" marker and sharing
> it's location with the sounding probe......or something along those
> lines. Hopefully someone else has the answer...
>
Once you have the probe, you can just enter in values for Lat/Lon in the
display which appears in the display tab in the dashboard, and the probe
will move to the point of your choosing.
Feature request/idea: The field selector Tom referenced allows for a
subset region of points, but does not allow for a single data point or a
series/list of coordinates. This would probably resolve Kevin's
issue and make it easier if one had many specific points to probe.
Cheers,
Brendon
--
*************************************************************
Don Murray UCAR Unidata Program
dmurray@xxxxxxxxxxxxxxxx P.O. Box 3000
(303) 497-8628 Boulder, CO 80307
http://www.unidata.ucar.edu/staff/donm
*************************************************************
