[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[IDV #IFD-863611]: Time-Height Display Question



Greetings Stefan,

> Sean,
> 
> Thanks for your thorough, albeit, disappointing responses. It is
> unfortunate that these features do not exist in IDV. It makes it very
> difficult to use IDV as a research visualization tool, when GrADS do
> all of the aforementioned with very little trouble.  Another feature
> IDV NEEDS is spatial averaging. I do not mean smoothing using weighted
> grid points, but area averaging similar to the functions ave and aave
> in GrADS (if you want more information on those functions go here: 
> http://www.iges.org/grads/gadoc/gradfuncaave.html)
> . Along these lines, there needs to be documentation on how all these
> different smoothers differ in the IDV. I know the 5 and 9 point
> smoothing routines, but I have no idea how a cressman weighted,
> circular aperture or the other smoothers work. Sure I can smooth my

I know we will talk next week, but I thought I'd pass along a few references 
regarding the Cressman (a circular radius of influence) and Barnes (a Gaussian 
radius of influence) routines you mentioned:

Cressman, G. P., 1959: An Operational Objective Analysis system. Mon. Wea. 
Rev., 87, 367–374.

Barnes, S. L., 1964: A technique for maximizing details in numerical weather 
map analysis. J. Appl. Meteor., 3, 396–409

Both are popular techniques for smoothing (and gridding observational data). 
While they do indeed make pretty pictures, they can cause a lot of trouble in 
spectral-space, so one has to be quite careful when using them. Trying to 
document this would/will be very difficult, as it took two months of a grad 
level objective analysis class just to cover these two techniques and how to 
properly use them.

See you next week,

Sean

> data, but every smoother has different intentions and different
> methodologies. The weights of the smoothing needs to be documented
> somewhere, otherwise I am creating images for publish without really
> knowing how I am smoothing (smoothing for the sake of smoothing).
> Furthermore, anyone trying to replicate my results might not be able
> to if I can't document how I created the images. Thanks again for your
> help, and will talk more next week.
> 
> V/R,
> 
> Stefan
> 
> Stefan Cecelski
> Ph.D. Candidate and Graduate Research Assistant
> AOSC Department: University of Maryland College Park
> 2343 Computer and Space Sciences Building(CSS #224)
> 703.801.1370
> address@hidden
> 
> 
> 
> 
> 
> On Oct 19, 2011, at 5:38 PM, Unidata IDV Support wrote:
> 
> > Greetings Stefan,
> >
> > The IDV cannot display the time-height plot as a time-pressure plot.
> > I may be going out on a limb here, but I think this is why. The user
> > guide shows the time-height functionality using wind profiler data,
> > which makes me think that this kind of plot was developed to show
> > wind profiler data. If that's true, then one would never need the
> > capibility to change units to pressure, as wind profilers do not
> > give us pressure. Since it turns out that time-height plots are also
> > cool for model output, the functionality was extended to grids. I
> > think we would need to add a time-pressure plot to do what you would
> > like to do.
> >
> > Plotting two variables on a time-height plot isn't possible int he
> > IDV either at this point. You could create multiple time height
> > plots, share their data probes, and view them side-by-side. The
> > other option would be to plot the variables on a data transect (in
> > which you could do a pressure coordinate for height) and animate in
> > time).
> >
> > As for your last question, there isn't any capability in the IDV to
> > smooth out probe time series data. However, you could create a
> > jython function to do such a thing, like a running time average.
> >
> > Sorry to be a downer with all of the "can't" responses. That said,
> > let's talk next week at the User Comm and see what we can work out.
> > I think there are solutions for at least two of your requests.
> >
> > Cheers!
> >
> > Sean
> >
> >> Full Name: Stefan Cecelski
> >> Email Address: address@hidden
> >> Organization: University of Maryland
> >> Package Version: 3.0u1 build date:2011-10-01 07:08 UTC
> >> Operating System: Mac OS X
> >> Hardware: Java: home: /System/Library/Java/JavaVirtualMachines/
> >> 1.6.0.jdk/Contents/Home version: 1.6.0_26 j3d:1.5.2 fcs (build4)
> >> Description of problem: Hello,
> >>
> >> I have a couple questions about the time-height view under Probe
> >> displays. First, is it possible to plot more than one variable on
> >> the same time-height display? I see you can change the parameter,
> >> but I cannot find how to add an additional parameter. Second, is it
> >> possible to have the y-axis as pressure levels rather than
> >> alititude (height in meters)? Pressure level are more useful for
> >> referencing work and are a standard for meteorological research.
> >> Finally, is there anyway to smooth probe data (like a 9-point
> >> smooth) instead of just using one exact point? This would be useful
> >> to have since when trying to find a circulation center (say, for a
> >> tropical cyclone), it might be difficult to locate a point center.
> >> Doing some sort of averaging or smoothing allows for some error
> >> while trying to locate the center, and will yield the
> >> characterisitics of interest to be still captured, even if the
> >> circulation center is the exact center.
> >>
> >> Thank you very much in advance!
> >>
> >> V/R,
> >>
> >> Stefan Cecelski
> >>
> >>
> >
> >
> > Ticket Details
> > ===================
> > Ticket ID: IFD-863611
> > Department: Support IDV
> > Priority: Normal
> > Status: Open
> >
> 
> 
> 


Ticket Details
===================
Ticket ID: IFD-863611
Department: Support IDV
Priority: Normal
Status: Open