[wxqc] Fwd: Weather data quality report for C6155 on 2008-12-15

Thomas F. Giella kn4lf1 at verizon.net
Tue Dec 16 08:05:21 CST 2008


I use accurate analog weather instruments to check accuracy of my Davis VP2. As far as barometric pressure I check it against my three local ASOS stations (KBOW, KGIF, KPCM) and it is always within .01".

The QC is useful but I take with it says with some caution. Your station could get flagged due to a particular local microclimate and/or another nearby station(s) could be inaccurate.

In a my opinion people worry to much about the QC data.

Take Care,
Thomas F. Giella, KN4LF
Lakeland, FL, USA
kn4lf at arrl.net

CWOP #AR692/KN4LF
CoCoRaHs #FL-PK-18

Lakeland, FL Daily Climatological Weather Data Archive: http://www.kn4lf.com/kn4lf22.htm 
Harmful Man Induced Climate Change (Global Warming) Refuted: http://www.kn4lf.com/globalwarminglie.htm

  ----- Original Message ----- 
  From: Victor Engel 
  To: Discussion of weather data quality issues 
  Sent: Monday, December 15, 2008 10:08 PM
  Subject: [wxqc] Fwd: Weather data quality report for C6155 on 2008-12-15


  I have a suggestion for these reports. This report for my station essentially is a red herring. But I understand why it happened. It's because both barometer and temperature spiked. And that brings me to my suggestion.

  It shouldn't be too hard to check for spikes in the data. When a spike occurs, there could be a temporal shift to do a curve fit. The temporal shift would correspond to a movement of a front.

  If you look at my graphs, they really match the analysis pretty closely. It's just that when the graph has a very steep curve, there can be an error simply because of the frequency and phase of the observation. By phase I mean that suppose I report every 10 minutes on the 10s and another station reports every 10 minutes on the 5s. I'd call that out of phase but at the same frequency.

  We could both report the same spike and it would be 5 minutes apart just by virtue of being out of phase with each other. The upper bound of such mismatches due to phase would be the frequency, 10 minutes.

  So how would the change be implemented? Probably the simplest would be to add code to identify a spike, and then, when an error is detected, try matching the data again after shifting it temporally by one unit of the sampling frequency. If there is a match, then don't record the error.

  On the other hand, it's no big deal if the source of the error is obvious.

  Victor

  ---------- Forwarded message ----------
  From: Weather Quality <weather-qc at nospam.gladstonefamily.net>
  Date: 16 Dec 2008 02:52:46 -0000
  Subject: Weather data quality report for C6155 on 2008-12-15
  To: brillig at gmail.com


        Date UTC Alt (mb) Temp °F DewPt °F Wind Dir Speed knts 
        15-DEC-2008 2315 1028.0 -2.00 34 -2.50 18 -2.39 5 -13 12.2 -6.0 
        15-DEC-2008 2325 1028.0 -2.00 34 -2.50 18 -2.39 19 -27 7.8 -1.6 
        15-DEC-2008 2345 1028.1 -2.10 34 -2.50 20 -3.69 342 +10 7.0 -0.8 
        15-DEC-2008 Errs 5/127  3/127  2/127  1/127  1/127  
        15-DEC-2008 Smry -0.7 0.80 -2 2.55 -2 2.26 +3 22 -1.0 3.1 

  Note that times are in UTC. The values displayed are 'Observed Error' The error value is 'analysis - observed'. I.e. if your observed value is higher than the computed value, then the error will be negative. The error readings are listed in bold. The row with the time of 'Smry' is a daily summary and the data is 'mean standard-deviation' for each observation during that day. The 'Errs' row contains the number of samples in error / the number of samples (as seen by MADIS). This may indicate more observations that are shown above. This is due to the differing limits used by this reporting program and the MADIS program itself. 

  For more information: 

    a.. Daily graph. 
    b.. Site status page including quality rating. 
    c.. Quality checking information. 
    d.. WXQC Mailing list. 
  To stop this email, just click here 




------------------------------------------------------------------------------


  _______________________________________________
  wxqc mailing list
  Post messages to wxqc at lists.gladstonefamily.net
  To unsubcribe or change delivery options, please go to:
  http://server.gladstonefamily.net/mailman/listinfo/wxqc
  To search the archives: http://www.google.com/coop/cse?cx=008314629403309390388%3Aknlfnptih9u

  The contents of this message are the responsibility of the author.


------------------------------------------------------------------------------



  No virus found in this incoming message.
  Checked by AVG - http://www.avg.com 
  Version: 8.0.176 / Virus Database: 270.9.18/1851 - Release Date: 12/16/2008 8:53 AM
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://server.gladstonefamily.net/pipermail/wxqc/attachments/20081216/638f25dc/attachment.html>
-------------- next part --------------

No virus found in this outgoing message.
Checked by AVG - http://www.avg.com 
Version: 8.0.176 / Virus Database: 270.9.18/1851 - Release Date: 12/16/2008 8:53 AM


More information about the wxqc mailing list