meteohub.de

This isn't a problem with Meteohub as such but I think it could be resolved in the software. About once or twice a day I get a solar radiation reading from my WH3080 that is much too high, typically the value reported is 1172 W/m2. This may be a problem with the sensor, which otherwise reports perfectly sensible results, but I had the thought that as most of the parameters measured by weather stations have practical limits would it be possible to filter these in the software rather than having to edit the data. For instance, my understanding is that a solar radiation level of more than 342 W/m2 is simply not possible and practical levels are highly unlikely to be greater than 200 W/m2. Is this a a possibility for a future update, simply recording all values in excess of 342 W/m2 as 0 on the basis that they must be incorrect? Presumably the same could be done for other parameters.

Cheers
Gordon

Gordan,

you seem to confuse average radiation with current levels of solar radiation.
the max value my Davis VP2 has recorded is
1389 W/m²

actual solar radiation hitting the ground depends on many things:
cloud coverage
sun height
day of year

http://en.wikipedia.org/wiki/Solar_radiation

And last:
Your WH.3080 basically measures and displays values in Lux and not Solar radiation ...

Yes, you may well be right regarding the average level issue but I think your Davis is not giving correct results either. I've done a bit more reading on the subject and find that the maximum solar radiation at a distance of one astronomical unit is generally taken to be about 1366 W/m2 BUT this value reduces at the earths surface due to atmospheric absorption and refection, going down to about 1000 W/m2 when the sun is dirctly overhead on a clear day. I've actually found a solar radiation calculation spreadsheet at http://www.ecy.wa.gov/programs/eap/models/solrad.zip and I've run the numbers for your location based on the lat, long and altitude you give on your website. The calculated maximum level (using 2 minute values, which tends to give the highest results) for August 2011 is around 830 W/m2, well below the 1389 W/m2 reported by your Davis, which would actually be above the accepted extra-terrestrial level. The predicted maximum value varies according to the algorithm used so the 830 W/m2 may well be on the high side.

I'm aware that the WH3080 actually reports Lux, I assume that Meteohub uses a conversion factor to get the value in W/m2 based on the spectrum of sunlight.

All a bit confusing but I can still see a place for a Meteohub software mod that filters obviously incorrect data.

Gordon

well,
those maximum values are measured not on summer days with no clouds in the sky, but on days with scattered clouds on which an effect called diffusion is increasing the radation level on a given point.
On cloudless days certainly I also have max. of 800-900 W/m² in summer.

Another myth is that one can calculate solar radiation (recorded by a pyranometer) from a luxmeter (a photometric unit) with an easy linear formula.
Without knowing the exact spectrum distribution of the light the conversion from lux into W/m² is impossible.

Here you see a chart showing the conversion factor K depending on the wave length distribution.
A conversion factor of K=100 is only correct on days with clear sky and a high standing sun.
All other cloud coverage changes K dramatically, so that folks who come up with a linear conversion formula can forget that.



Sorry to say so, but your weather station can measure and display Lux and Lux only.
Any conversion into solar radiation values without the chart above is absolutely incorrect and useless.

The solar radiation Meteohub displays for a WH-3080 is derived by a constant divisor from the Lux. This is just a rough estimate as wfpost has explained. However, I am sure it is not even close to being accurate, but better than nothing?

You can make use of the calibration function to tweak it a bit, so that it at least does not deliver too disturbing values at your physical location, or just ignore that value totally. The station itself just delivers uv index and lux.

admin wrote:The solar radiation Meteohub displays for a WH-3080 is derived by a constant divisor from the Lux. This is just a rough estimate as wfpost has explained. However, I am sure it is not even close to being accurate, but better than nothing?

You can make use of the calibration function to tweak it a bit, so that it at least does not deliver too disturbing values at your physical location, or just ignore that value totally. The station itself just delivers uv index and lux.

That's understandable, could you let me know the divisor so that I can use the calibration factor to convert it back to Lux? I could then change the units in the php files to display it correctly although I'm not sure if I can change the graphs to the correct units. Is there a format command in Gnuplot that I could use?

Thanks for the help.

Gordon

You have the conversion factor between Lux and Wqm already at hand when you simply devide the values you see on sensors, I guess it was 178 or so...

There are quite a few posts here that give examples how to define format strings. Please search for "gnuplot format" and you will get quite a few examples.