calibrating UV using multiplier

[jturner]

Hi all

I have been watching the UV readings for Canberra published by ARPANSA (http://www.arpansa.gov.au/uvindex/realtime/can_rt.htm) and comparing them with the UV readings from my weather station.

It seems like my weather station (FO WH-3081) gives a reading approximately three-fold higher than the value from ARPANSA.

Is there any simple way to include a multiplier, rather than an offset, for the UV reading? For example, can I add 'UVMult=0.333' to the 'Offset' section of the cumulus.ini file, or does it require something more complex?

Many thanks
Jeremy

[steve]

The Fine Offset UV reading does seem to give very high values, a number of others have commented on this. Geoff in Harwich said his seemed to be reading a fixed number higher than the presumed actual value, so I added the 'offset' calibration value. If you think a multiplier would be more appropriate, I can easily add one.

[jturner]

Hi Steve

Thanks for really quick reply.

If you could add a multiplier for the UV reading, that would be great. Once it's included, I'll let you know how effective it is for calibrating the UV readings from the Fine Offset station.

BTW, just wanted to once again say thanks for the amazing work you do with Cumulus. I have a young son (20 months old) and can't wait to teach him some of the fundementals related to weather and weather measurement and forecasting. He loves waking up every morning and checking the anemometer from his bedroom window. Your software will be fantastic to help me teach him to be excited about the world outside.

Cheers
Jeremy

[geoffp]

Hi Jeremy
My U.V reads about 4 to 5 higher than what is forecasted this seems to be a common fault with these sensors, I have tried replacing mine thinking that it maybe just a faulty sensor but it was still the same, like Steve said he put in an offset and I have a minus 4 set in there all of the time, it will be interesting to see what happens with a multiplier.

B.T.W love the bit about the youngest weather buff nothing like starting them young

[AllyCat]

Hi,

I was quite surprised to discover that the UV index uses a linear scale, so indeed a multiplier is the "correct" correction to apply.

The UV index requires that the sensitivity at 325 nm (i.e. "far UVA") should be only about 3% of that at 295 nm ("near UVB"). I suspect that FO haven't used a filter to provide this degree of selectivity, so the UV readings are never going to be generally "accurate".

To me, the only value of the UV sensor is that its scale fits more comfortably on a combined graph of temperature, etc. than Lux units.

Cheers, Alan.

[geoffp]

Thanks Steve for adding the multiplier for the UV it is now just trial and error until I get the right setting but even at the moment it is looking a lot more sensible

[jturner]

Hi Steve

Thanks for including the UV multiplier and the Light webtag in the latest build. They're working really well. I'm working on a multiplier that best approximates the UV reading, and at the moment it's about x0.33.

Just one question.....for the light value that's uploaded to the website, is there a way I can change the value from scientific notation to a whole number (eg 55673 Lux rather than 5.6E104)?

Cheers
Jeremy

[steve]

Just one question.....for the light value that's uploaded to the website, is there a way I can change the value from scientific notation to a whole number (eg 55673 Lux rather than 5.6E104)?

That was an error on my part which I've fixed for the next build.