Topic: UV index=cloud %

[Skywatch]

I have a UVN800 to an Oregon Scientific WMR100. I have the UV readings being graphed. How much info about clouds can you get from varying fluctuations in the UV readings?

[gadget_guy]

I'm guessing less than We'd like.  I was thinking the same thing today.  I know when I see spikes in the UV index that there was a passing cloud.  Alas, I don't think there is any way to extrapolate % of cloud cover.  When I don't see a high UV index over a long period I guess I can assume lots of cloud cover.  Other than that, I think it's hard to draw conclusions

[Bashy]

Hi

I use a modified OS UV to work with my WMR928, the UV coincides very well with the solar

Its been pretty cloudy here this morning, if you look at my graph (attached) you can see that
you can get a rough idea as to the cloud cover %

[KeithBC]

UV does respond to cloud cover, but it also responds to other factors, such as the composition of the upper atmosphere. 

Some time, when I have a bit more data to work with (I only started recording UV this summer), I would like to correlate UV and solar radiation, which does indicate cloud cover.  By removing the cloudy/sunny component of the UV readings, what is left should tell me what is happening in the upper atmosphere.

So many numbers; so little time.   

[Skywatch]

Thanx guys! The reason I started this topic is because some software programs and I see this on several Wunderground stations with UV sensors in the tabular display, sky conditions are made based on the rigidness on the UV graph. Now my station on Wunderground (KTXMCKIN27) also reports UV, but not sky conditions. I'd have a hard time gauging the frequnecy of the spikes to determine the cloud intensity. One other thing I have also asked reciently is how to take the UV index and convert those numbers into nanometer numbers or the Ultraviolet spectrum. I geuss UV sensors only go so far as sensitivity and scaling. Unlike Solar radiation (W/M2) which ranges from 0 to a couple thousand W/M2s.

[Scalphunter]

 You can still get sunburn from UV on an cloudy day. What i would do is what i done here from cloudy cover is an Thermometer(wireless) in an jar and then plot the reading from that. It is used here as an poor man cloud detector. By  setting up smart forecast I set my weather icons.


John

[Skywatch]

Well I have heard of the, "temperature in a jar" setup. But the question is, is it accurate? I mean the temperature in a jar, and the componets that actuall pyranometers and radiometers use are totally different.  My UV sensor is a UVN800 for my WMR100. Now when a cloud passes over the sensor, the UV has a down spike. The high UV for the day will show up as a flat line around 11:00AM to 2:00PM for that part of the day. Like for winter you can expect 3 on a sunny day. That's apart from somewhere around 7 and higher during spring, summer, and fall. And the number that the UV dips can give me the cloud thickness. Like if the clear sky UV is at a high of 5 and dips to 4 during certain parts and 3 and so forth on other parts of the day, but comes back up to 5 indicating the UV isn't going down, but being blocked by a cloud or even a bird, trust me thats happened. But this data on sky conditions doesn't show up on Wunderground. Like some other stations in the area that report UV and not solar radiation will have in the tabular little sky condition abreviations like CLR., SCT., BKN. I think It might be the way by software processes and reports the data. But some software programs will base sky conditions off of UV or any other sun data. I will look into this further. 

[Scalphunter]

 Accurate enough that I can set clear sky at night as it will drop 3-5 degrees below OAT as it also sit on black bottom. Go ahead and  doubt me  but if you don't try it it be your loss.

 It  shows here on my graphs each andevery cloud that goes by and sharp drop off at sunset.

John

[Skywatch]

Well I have figured it out. A down spike indicated a cloud, now the number the UV# goes down can tell me the thickness of the cloud. The size of that spike, like from when it drops to, when it comes back up will tell me the size of cloud which is assisted by wind speed measurements. And the frequency of those fluctuations will tell me the sky percentage and coverage.

[gadget_guy]

I have the UVN800 and I'd be hard pressed to infer too much from the UV index alone.  I don't think it makes for an accurate reading of the percent of cloud cover or can accurately substitute for a good visual observation.  I think you can make some inferences i.e.  Very cloudy, or broken scattered clouds depending on the UV index value and the frequency of downward spikes, but I don't think you can estimate cloud cover from the readings even when combined with wind speed etc.  Just saying.

[NeverDie]

Hi

I use a modified OS UV to work with my WMR928, the UV coincides very well with the solar

Its been pretty cloudy here this morning, if you look at my graph (attached) you can see that
you can get a rough idea as to the cloud cover %

How are you acquiring the solar data?

[mcrossley]

My UV readings also sometimes spike upwards with clouds - they reflect some UV downwards, especially when the Sun is lower. How would you factor that in? I really don't think you can infer cloud cover with any accuracy using the UV reading on its own.

[wxtech]

On my web site, I report sky conditions at several regional airports.  I don't have a sky condition function on my wx station so I use the cloud base calculation. In VWS it is cloud base (ft) = 250(Temp - dew point).  Another formula = (temp - dew point)/4.4 times 1000.  Use 2.5 for °C.
My calculated cloud base follows closely with the regional sky conditions in this table: http://www.lexingtonwx.com/#metar
This calculated sky cover indication works at night also when the UV & solar indications are zero.  My system is a Davis VP2+ with solar and UV sensors.

[Skywatch]

Forgot about this thread.

Love Weather Display because it calculates cloud cover based on cloud cover. Since I began this thread I got rid of the WMR100 and 200. One thing I found out is the UV to solar isn't too good. The solar % would always read 100 percent. The W/M2 would read in 100 W/M2 per 1 UV index value. Inaccurate. Also now that I've upgraded my VP2 with solar and UV sensors I've noticed at any time of the year the solar and UV don't agree. The solar and UV seem to do their own thing.