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

Re: 2000118: 20000118: unit errors in decoded sounding data (fwd)

Hi, Chiz - in the title below, "Chris" in Albany was not reading from my
site, rather he was reading from the SUNYA site - I don't have any such
labelling in my plots or tabulations of data - I run gempak and put the
raw output out on my CUDOS site unfiltered.  Chris was referring to errors
I think in the formula implementation speed = sqrt(u**2 +
v**2) (regardless of units).  

If you do calculations using his examples of u=1.6 and v=-1.3, you don't
get 4.0 due to round off error.  Albany is massaging some data here to
simplify their data listings, but we don't.  In other words, this is not
an issue for support after all, sorry for the trouble!  I've copied Albany
on this, too.  Paul

On Tue, 18 Jan 2000, Unidata Support wrote:

> 
> Paul,
> 
> I was refering to you title:
> >> >PRESS     WDIR    WSPD    UWND    VWND
> >> > MB        DEG    KNTS    KNTS    KNTS
> 
> In particular, if you are plotting UWND and VWND in GEMPAK, then
> the units are m/s not knts as you have labled.
> 
> If you want to use knts in GEMPAK, then use UKNT and VKNT.
> 
> I was assuming you were refering to data as listed by GEMPAK, though
> you did not refer to using GEMPAK. Since the 4 characater
> parameter names are simmilar to GEMPAK, but certainly not "PRESS"-
> I believed you were refering to snlist with some cooked up
> column lables from your own since that listing you
> provided is definitely not GEMPAK.
> 
> Otherwise, I have no idea what you mean in your response below.
> 
> Steve Chiswell
> Unidata User Support
> 
> 
> 
> 
> 
> >From: Paul Ruscher <address@hidden>
> >Organization: .
> >Keywords: 200001182100.OAA16583
> 
> >Thanks, doh!  My bad - I think he got this example from Albany - in any
> >case, it is round off error, of course!  My site includes data to two
> >digits and does not have this problem.  1 digit rounding for both u and v
> >can lead to +/1 5% error or so, which is enough to explain his
> >problem.  I'll write him again!
> >
> >Paul
> >
> >On Tue, 18 Jan 2000, Unidata Support wrote:
> >
> >> 
> >> Paul,
> >> 
> >> UWND/VWND is meters per second. UKNT/VKNT are knots. If you want units
> >> in knots, you need to use the appropriate parameter.
> >> 
> >> The units column label is not from GEMPAK, but apparently something you
> >> have tacked on. You need to maintain that label to be consistent
> >> with the parameter usage above.
> >> 
> >> Steve Chiswell
> >> Unidata User Support
> >> 
> >> 
> >> 
> >> >From: Paul Ruscher <address@hidden>
> >> >Organization: .
> >> >Keywords: 200001181842.LAA08774
> >> 
> >> >Gempak decoding problem here, perhaps?  I've checked and I think Chris is
> >> >correct - happens in other soundings, too.  Paul Ruscher
> >> >
> >> >---------- Forwarded message ----------
> >> >Date: Tue, 18 Jan 2000 15:10:32 +0000 (GMT)
> >> >From: Chris Walcek <address@hidden>
> >> >To: address@hidden, address@hidden
> >> >Subject: unit errors in decoded sounding data
> >> >
> >> >To whom it may concern,
> >> >
> >> >I regularly download and use decoded soundings
> >> >from your web pages. There is an error in the
> >> >units of the wind speeds presented in these tabulated
> >> >data. For example, here is a excerp of Tuscon AZ
> >> >0Z sounding, 18 January 2000
> >> >
> >> >PRESS     WDIR    WSPD    UWND    VWND
> >> > MB        DEG    KNTS    KNTS    KNTS
> >> >-----    -----    ----    -----   ----- 
> >> >929.0    310.0     4.0    1.6    -1.3 
> >> >925.0    310.0     4.0    1.6    -1.3
> >> >
> >> >
> >> >First, wind directions and speeds are presented
> >> >in degrees and "KNTS". Speed I assume is in
> >> >nautical miles per hour.
> >> >
> >> >Later, U(west-east) and V(south-north) winds
> >> >are tabulated, also in units of KNTS. However,
> >> >when I take the square root of the sum of
> >> >the squares of U and V (sqrt(U^2 + V^2) one
> >> >does not get the original wind speed. In fact
> >> >there is a discrepancy of a factor of 1.94 which
> >> >is precisely the conversion of nautical miles per
> >> >hour to meters per second.
> >> >
> >> >It looks like the units of U and V are WRONG and
> >> >should be in meters per second (m/s)????
> >> >
> >> >Is this correct? If so please change the table
> >> >headings appropriately. Other explanations
> >> >of this discrepancy are possible.
> >> >
> >> >Thanks
> >> >Chris Walcek
> >> >
> >> >(beware the Mars pathfinder fiasco where 
> >> >sloppy unit conversions were responsbile for
> >> >disaster!!)
> >> >
> >> >
> >> 
> >> ****************************************************************************
> >  <
> >> Unidata User Support                                    UCAR Unidata 
> >> Program
> >  <
> >> (303)497-8644                                                  P.O. Box 
> >> 3000
> >  <
> >> address@hidden                                   Boulder, CO 80307
> >  <
> >> ----------------------------------------------------------------------------
> >  <
> >> Unidata WWW Service                        http://www.unidata.ucar.edu/    
> >>  
> >  <
> >> ****************************************************************************
> >  <
> >> 
> >
> 
> **************************************************************************** <
> Unidata User Support                                    UCAR Unidata Program <
> (303)497-8644                                                  P.O. Box 3000 <
> address@hidden                                   Boulder, CO 80307 <
> ---------------------------------------------------------------------------- <
> Unidata WWW Service                        http://www.unidata.ucar.edu/      <
> **************************************************************************** <
>