WS3083 Light sensor

[sarakiniko]

I have a newly set up WS3083 It all seems to work well with Cumulus. However it is recording up to 22hours of sunshine. I would love this to be correct but Devon is not that sunny in January or even June for that matter. Any suggestions?

[steve]

You may need to adjust the settings in the 'solar' section in the Cumulus station settings. There is information in the help, and there may also be something in the wiki.

But to be recording 22 hours of sunshine a day at this time of year sounds like you may have a faulty sensor, as it should be recording zero while it's dark, and with the default settings Cumulus doesn't record that as sunshine. What does your solar graph look like?

[sarakiniko]

Thanks for your response. The solar readings in cumulus are responding to the weather station output. The attached file shows this but I don't know how to change the solar parameters to reflect the readings as sunshine hours. I have now changed the Threshold to 50% and min to 5 W/m but that presumably will not affect the history.

[steve]

It can take a bit of tweaking, particularly with a Fine Offset station which isn't even measuring W/m2 anyway. Someone with a similar station can hopefully offer you their settings as a starting point. It's still a bit odd that you're getting 22 hours recorded, as the reading does seem to be dropping to zero at night. That min value setting should fix that.

[mcrossley]

Your detection threshold of 5 W/m2 is very low, the WMO recommend 120 (some other organisations use 100 iirc), though I actually use 50(!), I do NOT have a Fine Offset station though. The lower the figure the more sunshine you will register.

Your theoretical figures also look a little on the low side, have you tweaked the default parameters?

[steve]

Are you sure about that, Mark? I'm using 9 W/m2 as my minimum value, and that seems to work quite well for me. I definitely have sunshine with readings in the low teens. If I used 100, I would hardly record any sunshine at all in winter. My theoretical max at solar noon is 102 at the moment.

[mcrossley]

Yes, it's surprisingly high. Burt talks about in his book, how they stopped using one electronic sun sensor because it had a threshold of 87 (30% below the WMO threshold). The problem is historic sunshine figures are based on Cambell-Stokes figures, and you need a fair intensity of sunlight to get the paper burn - far higher than modern electronic detectors are capable of. So in order to make the stats comparable, a relatively high threshold has to be used.

[steve]

Ah, OK, so the difference between sunshine and "bright sunshine"? I think I'll stick with what I've got, I'm confident that most people would regard the sun as shining when my settings say it is.

[mcrossley]

I agree, that is why I use a lower figure too, this is for my amusement, not to feed into official stats.

But at 5, you have very little wriggle room regarding the difference between measured and theoretical, especially with a FO which hasn't the most sophistcated of sensors.

[nitrx]

For my WS3080 I use a treshold of 63 % and a minimum of 15 W m2 in cumulus.ini I've set LuxToWM2=0.0079 http://wiki.sandaysoft.com/a/Cumulus.ini#Section:_Solar

These values gives me raisonable figures , but I'm sure you have to 'play' somewhat with the values because the Lux sensor isn't a real solarsensor ofcourse.

[sarakiniko]

Thanks for the comments and suggestions. I've modified the parameters again to Threshold 70% and Min to 20 W/m2. The sun was definitely bright this am when the Solar reading was about 40 W/m2.

I am still unclear about the theoretical radiation value. Is there any way I can determine what it should be? Maybe I need to talk to somebody down the road at the Met office but I fear they might get too technical for me! Is it appropriate to fiddle with the Trans factor?

Nigel

[steve]

The transmission factor is an indication of how much is lost through the atmosphere where you are. It's there to be adjusted if it helps. This isn't an exact science; the way the Davis sensor works isn't really suitable for determining sunshine hours accurately, so we're all having to do the best we can. The Fine Offset sensor even less so.

The way I adjusted my transmission factor - I waited for a very clear, cloudless day (probably in summer, but I'm not sure that matters) and then adjusted the transmission factor so that the theoretical calculation came out at about the same as (or very slightly higher than) my readings, on the basis that I must be getting close to the maximum on a clear, cloudless day.

[mcrossley]

What Steve said, plus be aware that your actual reading can be higher than the theoretical if there are broken clouds around (they reflect more light onto your detector). So when doing any calibration is is important that it is cloudless.