Problem with uv and solar time

hi Brian

could you please have a look at
the uv and solar data were correct, [color=red]as soon as we changed to the UK summertime by adding 1 hour

The time in the graphs are showing maximum UV and solar for the day is at one o’clock, [/color]

Take off the hour and the graphs come correct, but obviously this affects all the other data

I am using a one wire solar sensor
uv sensor mod for the WMR928NX

mick

I to have this issue using modified UV for WMR928 as shown in the image below,

To be honest i aint sure myself which one is 1 hour out, is it the UV or is it the Solar?
Anyway, this happend when the clocks changed an hour…


Untitled.png

I’m confused. Are you guys saying the peak solar and UV on the graph are at 1:00 PM “graphtime”?

But 1pm BST is midday GMT (UTC), which is the time of peak solar & UV in the UK. So your graphs are reading correctly.
The time at which the Sun reaches its highest point in the sky doesn’t change just because we go to BST, you would need to push the Earth’s rotation forward one hour to do that. :wink:

Martin: There have been previous posts like this thread where Simon appears to be saying the UV and solar peak at different times so I’m not really sure what the current issue is.

Hi guys

Yeah Niko is correct, the problem appears to be that my UV and solar high noon are
2 hours a part, you can see it more clearly in the image below, i would have thought
the high’s would have been near on the same, but it looks like the solar is around 1:30
to 2pm and UV is around 12pm, I thought they would be about the same?

Also i thought the worst times for sunburn were between 11am and 3pm but as you can
see from my graph its already dropping big time by 3pm, this is why i am thinking the UV
should be forward 1 hour on the graph?


Untitled.png

I think this has to be a hardware/optical issue with your detector. Maybe the actual UV sensor element inside isn’t horizontal and the sun being higher in the sky now has made the error worse. I really don’t see anyway that WD can be recording the data at the wrong time.

Hi Niko

The image added below i thin was taken before the clocks did the hour thing as you can see the
UV and solar are pretty much the same (solar noon and UV noon time) but when the clocks did the
forward and hour thats when this issue occurred…


No further thoughts on this?

I have the same thought as before, WD is just reporting what it is getting from the sensor, and my conclusion is that the optics of the sensor are “off”. Here are the facts as I see them:

  • I agree that the trailing/falling edge of the UV curve on the “bad” graph doesn’t follow the solar very well, but the leading/rising curve does
  • There’s a big difference in solar elevation between those two graphs, they both appear to be relatively sunny days but the solar on the “bad” one is twice the value on the March one
  • We discussed before that when you looked at the UV log file you saw that the most recent record was showing with the correct current time
  • Not tried before, but if you mouse over the graph on the screen do you see the UV values? If so check some against the logfiles
  • I drew a noon line on the “bad” graph. Your theory is that the point on the UV curve indicated by the blue arrow is really data from 1:00 PM and not noon - right? If that were true where does the value indicated by the red arrow come from? According to the graph it’s from 6:00 PM, but according to your theory it would have to be from 7:00 PM, a time that hasn’t happened yet

Sorry, I just don’t buy the time warp thing :frowning:


bashuv.gif

I know what your saying but how come it was fine before the clocks changed?
This is the part thats baffling me.

All i can tell you is this,

  • The sensor is pointing south, but that dont matter it could point north if i wished
  • The sensor is horizontally plumb (most important i think)
  • There is nothing in the path of the sun to the sensor during the daytime
  • There is obstruction (houses) during dawn and dusk, this is why the trailing edges are how they are
  • The sensor has not been moved from its position since about 8 months ago

The solar would be considerably higher now than in march

Please do me a favour and see my graphs here http://www.northantsweather.com/wxgraphs.php
Scroll down to the calendar and go to march, the best day before the clocks changed is the 21st
you can see that the morning wasnt fully sunny but it got there just before lunch, notice both the
solar and UV are pretty much in line at their highest points.

Now please go to the last Sunday (thats the day the clocks did the change and you will see a big
difference!

I dont know how, i dont know why, all i know is its doing it and it wasnt before the clocks changed!

Looking again at the graph Niko highlighted, follow the dips in both UV & solar (clouds obscure the Sun) and you’ll see that they match, another indication that the timing isn’t the issue. :wink:

It’s not that your sensor is overheating and causing a bad reading?
If you look at your graphs for the early part of June this year, when it was cloudy, then the solar & UV tend to agree. It’s only when you have long periods of Sun light in the morning that the UV seems to tail off.

Bashy - have you considered that the data may in fact be real, that it is the conditions in the sky that are relevant to diffracting the much shorter wavelength UV which is more sensitive to haze and summer smog which would occur from about midday onwards and more so after DLS applied. What I am seeing in the “Bad” graph is just this. As the timing of the rise and end of day is still correct. You will need to compare against another UV sensor graph in your region or at the same latitude.

Graeme

Those are all good thoughts too :slight_smile:

I take all your points on board, but Mick has the same issue and we are both using the Modified UV settings!
I would also go with any one of your ideas and stick with it if it wasnt for the fact that it happened
straight after the clocks changed 1 hour, if it was any of your ideas/theories then it would happen
gradually over time, and not straight after the clocks change.

Also pointing out its 1 hour out on the graph since the clocks changed 1 hour forward, the solar on the graph
moved forward 1 hour but the UV did not follow this when the clocks changed it stayed the same as though is
still in BWT is that real? British Winter Time anyway lol

Ok to sum up cause i think its getting too complicated and too much being read in to it…

1). I am using the modified UV settings that Brian added not too long back for the WMR 928
Mick (the Orignal Poster) has this exact same issue using the exact same UV setup "uv sensor mod for the WMR928NX and modified settings for the setup

2). Before the clocks changed its was 100% OK, After the clocks went forward, the Solar followed suit but the UV did not it stayed set at the winter time

So, the UV on the graph needs to move forward one hour to line up with the solar that changed when the clocks went forward, for some reason the UV did not move forward on the graph. The UV data is correct its just one hour behind. ok so i repeated myself but i dont know what else to say :?

The Modified UV settings for the OS WMR 928 was a late addition that was added beginning of summer last year and not the same setup as the WMR200 even though its the same sensor.

I would be very interested in seeing anyone else’s graph that is using the modified UV settings for
the WMR 918/28/68 and that is the UK/Europe, but so far theres 2 of us that have this issue, if it was
just me then i would say its me thats broke something but as someone else is using the same setup and
settings has the same problem then i would say the probability of it being my error has dropped considerably

Bashy - have a close look at your recent full graph, or even the one at mid afternoon - can’t you see the graphs UV and Solar share exactly the same time frame - when a cloud passed over, both graphs dipped at exactly the same time - and this happened twice, ie they are perfectly in sync. The other thing with solar, is that the reflectivity from the clouds amplifies the solar signal even beyond 100% yet will strongly lower the UV signal which is what we are seeing in the two graph lines. There may be high level cirrus haze that you can’t see affecting the UV signal.

I believe what you are seeing is a perfect representation of the differences between UV and solar - they will behave differently due to the amtospheric conditions affecting their respective transmissivity through the atmosphere especially in the afternoon.

Sure, I did. UK changed time on March 29th.

Here are the graphs for March 30th and 31st, after the change. To me they look pretty well synchronized, if there was a one hour offset it would surely be seen in these two datasets.


mar30.gif

mar31.gif

I give up but heres a link to Micks graph http://www.weather-above.com/curr24hourgraph.gif

But if you find a WD main screen graph with UV and solar that shows the same as mine and
there not using the modified UV settings please show me, cause i done a good search this
morning on google and couldnt find any graph showing this issue other than Micks thats using the
same settings…

I cant wait till October…

That’s in perfect sync, there is nothing wrong with that. (although there is a low period for UV mid morning unexplained) You have an aerosol/particulate problem in your sky that is reducing the UV intensity and reflecting it or absorbing it reducing the intensity measured on earth. You have to learn about and know what causes UV to reduce in value due to atmospheric conditions, including the ozone layer… compared with Solar radiation. The UV is much shorter wavelength and is easier to relect than the longer wavelength of visible light - that solar sensors measure. The very short UV light is harmful to the unprotected skin and the eyes and is why it is measured using the UV Index.

Just a thought - have you tried resetting your PC/Station time to GMT for a couple of days and observing the results?