the docs of the sensor tells us that it measures both uva and uvb. it returns values on a scale from 0 to 11. It has a graph in figure 3 with the uv response. It has a linear scale. This looks close to the blue line in the second figure of the wikipedia article you referred to (note that that graph is on a log scale, so the values on the tail to the right are more then 10 times lower then the maximum, which matches the graph in the docs of the sensor).
So I think the intention is to return a value that IS related to sunburn and the UV index you see in other weather reports.
Thank you for your reply. My intervention to this thread is mainly because, in practice, I’ve not been able to correlate the WF UV figures with the official ones. I hope the following figs will give you an idea of my concern.
Although it’s not very clear, the deep violet line at the bottom shows the measured UV in WF, showing a UV index peak of 5.3 at 11:37 and generally following the solar light curve (orange) including the peaks and valleys.
This shows the official measured UV covering the same period with an index peak of 3.9. You may wonder why the curve is relatively smooth, compared with WF; this is because light cloud, visible in the WD graph, is transparent to UVB, less so to UVA and even less to visible. The intermittency of visible light in my graph is reflected in the UV curve of WF, whereas there is relatively little intermittency in the official UV one, related to visible light.
I’ll continue to study this and I’ll try to obtain a better UV/solar graphs to explain my concerns better.
2 Likes
That’s pretty interesting. I didn’t know that uvb passes through clouds. With what instrument is the official uv measured in your graph?
Yes, it passes through light clouds, which is why you can still sunburn with a hazy sun. It depends on the size of the water droplets.
The original instrument was the British Dobson spectrophotometer ( https://en.wikipedia.org/wiki/Dobson_ozone_spectrophotometer ), which had two sensors, one in near UVA and the other in UVB and did a comparison between them (The first one was installed in Davos, Switzerland in the 1920s). Although still much used, it has been largely replaced by the Canadian Brewer spectrophotometer which is much more sophisticated ( http://kippzonen-brewer.com/about-brewer/brewer/ ) and measures more variables. The local one is a Brewer, I think installed at the Institute of Meteorology of the University of Cyprus.
But I thought it was uva that passes through clouds due to its longer wavelength.
If you have fluffy clouds uvb can increase because it bounces off the clouds.
You are probably right that it depends on droplet size. I didn’t look into it.
To be honest, I don’t have much of a clue, weather-wise. My expertise (23+ years working with UNEP) is more in the effects of stratospheric ozone depletion at the earth’s surface and, above all, the emissions of ozone-depleting chemicals. Secondarily from this experience, I picked up some semi-scientific knowledge concerning measurements of the ozone layer and consequently UV radiation. That was in a former life over 20 years ago and, since then, I’ve forgotten more than I ever learnt!
https://uk-air.defra.gov.uk/research/ozone-uv/the-uv-index
Practically, yesterday was overcast and no comparison was possible. Today, it is mostly sunny with variable intensity and again I’m seeing the UV curve from WF closely following the solar in shape. Incidentally, the peak level (UV and solar) today is lower than 2 days ago, which I think is due to Sahara sand blowing across the island. I guess I should wait for a day with wall-to-wall sunshine to get a better idea.
1 Like
In the graph above, we have the solar radiation (yellow). In violet, we have the UV from the WF with a multiplier of 175 which arbitrarily takes it onto the same scale for the graph. In my experience, my guess is that the UV curve is from the values of near-UVA by virtue of its close resemblance to the solar curve. I would expect a UVB curve to be much more distinct from the solar curve. Unfortunately, I couldn’t find an official UV curve for that timescale, yesterday.
as mentioned before, in the docs it tells you that senser returns uv-index, which is based on UVA and UVB multiplied by some predefined sensitivity curve
I’m sorry, but I cannot agree with you. IMHO, the UV curve is essentially close UVA as I’ve demonstrated above with the curves. In fact, it is so close to the solar curve as to be almost indistinguishable.
1 Like
The Broadcom APDS-9200 device provides ultra-violet (UV-A and UV-B) sensing and ambient light sensing in a specially designed matrix arrangement for optimization. This allows the device to have optimal angular response for ultra-violet and ambient light sensing. The APDS-9200 converts UV light into digital data and display as UV index (1… >11) with higher values representing higher UV exposures. The ultra-violet sensing feature is useful in consumer applications for monitoring of UV-A and UV-B radiation as UV radiation is part of the electromagnetic spectrum that reaches the earth from the sun. APDS-9200 is able to measure UV wavelength which has been classified into UV-A (320 nm – 400 nm) and UV-B (290 nm – 320 nm). The ambient light sensing is targeted for display management with the purpose of extending battery life and offers optimum viewing in diverse lightning conditions.
perhaps you should not compare it to a curve of solar radiation, but to reported and trusted uv index values around you. Those values are also a mix of uva and uvb. they are mixed using some predefined sensitivity curve. (sensitivity of human skin for sun burn)
Now it might be true (I don’t know) that the sensor gives different values, as the sensor is not directly exposed to sunlight, but through some diffuse plastic window. That might block out some of the uva and uvb so that the ratio between uva and uvb changes a bit. If that is enough to report inaccurate uv index values, I don’t know. Are yours inaccurate?
Thank you for your reply. I agree with you that there are quite a large number of variables and it is impossible to be sure what is what. Unfortunately, the one and only trusted UV index on this island has been out of commission since last month, so I have no absolute comparison available.
The above graph is taken from the datasheet for the sensor; the bits in red have been added by myself.It is important to consider that we are looking at the energy required to produce the effects on the skin and the energy is a function of the surface area under the spectral curve. It is easy to see that the energy to the right of my red line is much higher than to the left.
“The differential effects of various wavelengths of light on the human cornea and skin are sometimes called the “erythemal action spectrum.”.[[50]]The action spectrum shows that UVA does not cause immediate reaction, but rather UV begins to cause photokeratitis and skin redness (with Caucasians more sensitive) at wavelengths starting near the beginning of the UVB band at 315 nm, and rapidly increasing to 300 nm.” (Wikipedia) This implies that the effect on the skin is proportional only to the area more or less to the left of my red line.
As you correctly say, the UV reaching the sensor may not necessarily represent the UV at the top of SKY, although one would hope that the difference is small. Notwithstanding, I respectfully suggest that the UV index, as displayed, is relatively meaningless in terms of the potential effects on the skin and eyes of the human body.
I apologise for appearing to be a pain in the posterior but I do know from my experience that the dermal effects are more affected by the thickness of the ozone layer, rather than the measurements of mixed UVA/UVB at the Earth’s surface.
1 Like
that is kind of strange statement. For sure dermal effects are directly influenced by uva/uvb reaching you! (and not directly by the thickness of the ozon layer). So measuring that makes total sense. The uva/uvb is multiplied by some standard sensitivity curve. That is just a crude thing to do, as the sensitivity varies a lot per person, but at least if gives people some indication of how likely sun burn is going to happen.
On the other hand the uva/uvb reaching the earth’s surface depends on the ozon layer, so sunburn increases indirectly by the thickness, but saying that it depends more on the thickness of ozon layer then uva/uvb is a strange thing to say.
If you want to check if the uv-index given by the device matches the real uv index, you need to compare it with trusted source. As you don’t have any on your island, you might just look at other weatherflow sky units and see how well they match the reported uv-index to a trusted source in their neighborhood. That might convince you (or not, depending on the outcome) that in general weatherflow reports accurate uv-index values.
1 Like
btw if you are just trying to say that sun burn is more effected by uvb then uva, then you are right, but that is taken care of in the sensitivity curve. In the curve, uvb counts 100-1000 times stronger then uva. So that reflects in the final uv-index that the sky unit reports.
Okay, I’ll try and explain where I’m coming from in simple terms. UVA causes ageing of the skin. UVB causes cancer and glaucoma. The SKY unit measures both UVA and UVB, but with a predominance of UVA. As such, because it does not give separate measurements of UVA and UVB, its index is weighted heavily to UVA by virtue of its response curve. It would therefore seem to my feeble mind that it is predominantly measuring the energy of the UVA and very little of the dangerous UVB. This is confirmed to a large extent by the shape of the UV curve in my graph above, which follows the visible light curve in shape to a remarkable degree. I would expect a UVB curve or one with the predominant ratio of UVB:UVA to be quite different as shown by the graph in my post of 24 January.
Perhaps the most important point is that UVB has a very different curve to UVA during a specific time period because of the chemical reactions that take place in the ozone layer, almost from minute to minute.
I’m not trying to say that the UV index as measured by SKY is useless. What I am trying to say is that it requires interpretation as a measure of the risk of overexposure to radiation, knowing that the average man in the street would not have a clue as to its real meaning in terms of long-term health.
You may be interested to know that my weather forecasts do take into account the stratospheric ozone in Dobson units (www.ternis.nl)
2 Likes
The Sky is not a scientific instrument. It is made for the consumer market.
Indeed. In fact, I’d argue for consumer purposes that either of the 3 light values is equally useful, as their response curves are extremely similar. Taken to it’s logical extreme that means that you are more likely to get a sunburn when the sun is really bright, which is a bit of a tautology, though it may not require much of a sensor. 
that is probably a typo, server not found
www.temis.nl does exist though
but anyway the resulting value does count way more stronger the uvb then the uva values.
If you look at https://www.rivm.nl/zonkracht/zonkracht-2016 you also see that it isn’t supposed to be a smooth curve.
yeah, but that is kind of true for most weather values. once you are outside, who needs a sensor 
anyway it is as it is. We have a sensor that measures uva/uvb and returns the uv-index. That uv-index probably resembles other uv-indexes reported from trusted sources. If that is useful for you or not is up to you.
1 Like