[wxqc] Has MADIS gone MAD ?

[Keith Miller]

> -----Original Message-----
> From: wxqc-bounces at lists.gladstonefamily.net
> [mailto:wxqc-bounces at lists.gladstonefamily.net]On Behalf Of Philip
> Gladstone
> Sent: Friday, November 24, 2006 10:44 PM
> To: Discussion of weather data quality issues
> Subject: Re: [wxqc] Has MADIS gone MAD ?
>
>
> It turns out that (in this case), something happened on november
19th
> for this station, and it's barometer readings went wrong.
>
> The thumbs up indicates that (over the last 4 weeks) it has been
> reasonably well behaved.
>
> I'm thinking that if the rating for the last 20% of the
> analysis period (default 4 weeks) is different from the rating
> for the whole period, then I should display the most recent
> rating, along with an annotation that indicates that it is a
> 'recent' change.
>
> Any thoughts?
>

What about a weighted scale, that is recent deviations hold
more weight. That way if someone has a "bad day" or two and
fixes the problem, they could get back to good QC quicker.
I would also show a recent problem quicker.


Keith
--
CW5250 => http://weather.stadhaugh.com