Hi Greg

Many thanks for your reply, you raise some very valid points about data in
the real world. Maybe I'll just leave it as it is :)

Cheers
Colin

On Sun, 13 Aug 2023 at 11:16, Greg Troxel <[email protected]> wrote:

> "[email protected]" <[email protected]> writes:
>
> > Hi all
> > I have a situation whereby (I think) I have some odd building
> reflections
> > spiking my solar radiation sensor. This results in occasional readings
> way
> > above the expected maxSolarRad value.
> > I'm going to try and sort out exactly what's happening but in the
> meantime
> > I wondered if it would be possible to use StdCalibrate to correct those
> > errant readings?
> > Putting it simplistically what I'm aiming for is a statement along the
> > lines of;
> >
> > radiation = maxSolarRad if radiation > maxSolarRad
> >
> > I'm not sure if that syntax is correct above or will even work,  but
> this
> > will hopefully remove any spikes greater than the expected maxSolarRad?
> >
> > Am I barking up completely the wrong tree?
>
> Yes.  You have data and you should record it.
>
> There are three situations, at least, where you will measure a radiation
> that is higher than maxSolarRad.
>
>   - small amounts, because the formula is approximate
>
>   - reflections from mirror-ish surfaces resulting in direct plus extra
>
>   - lensing from cloud edges, where you see = in clear sky, lower in
>     cloud, and a messy pattern as the edge happens
>
> In these cases, you really can get measurements that are valid and above
> maxSolarRad.
>
> The question is: what are you trying to do and what do you think it
> means?
>
> If you are trying to find some value that represents the highest value
> you saw more than momementarily, compute the 98th percentile of
> radiation, or 95th or 99th.  Then the data means what it says.
>
> As soon as you extract data with a conditional on theory, you have
> something that is very hard to reason about and describe.  My advice is
> don't go there.
>
> This is different from rejecting values of 120% humidity reveived from a
> sensor as obviously bit errors.  Seeing too-high radiation happens
> without corruption.
>
>
> I just record it, and graph it, and I don't worry that it's sometimes
> higher than theory.  Real data is like that.
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/weewx-user/CACjxfUsz36A026ft9YyH25SKNOuRYL4NCCUJPdkyMqHoQMB5pw%40mail.gmail.com.

Reply via email to