Cumulus Support

I have recently purchased a solar sensor to add to my VP2 setup and am still 'playing' trying to get the optimum set parameters within Cumulus, particularly as there has been little opportunity with a clear sky. I initially had the Trans factor set to .91 to keep my values within the curve, but from what I have read now believe some of my high values may be due to the effect of clouds and have started to lower this setting. Is there a value which would apply for Yorkshire or is it trial and error and once set, is there a way of applying this to previous days records ?

Thanks

It's a case of trial and error to find the right values for your location, but note that you're never going to get it perfect, the Davis solar sensor isn't designed that way. And yes, you will get occasional readings over the theoretical max due to cloud reflections etc. If you want to change data retrospectively, you will have to edit the data files.

I found that you need an almost perfect cloudless day to calibrate and check, then you need to judge how 'hazy' it is, it is rare at my location to get a cloudless day without haze. I think I have it just about right now, most cloudless days registering in the high 80 to mid 90 percents, the one really clear sky I witnessed recently hovering around 100%.

Clouds can make quite a big difference (+20%) to my non-Davis sensor, particularly at lower sun angles, but also if the reflection is right at higher angles too.

I get quite a few very clear cloudless days with no haze, and I find the theoretical curve isn't quite correct. It underestimates the radiation levels for sun elevations away from the zenith... or perhaps it overestimates them for high sun angles. I can set a transmission factor to get a best fit to the curves for morning and afternoon, but it doesnt match at solar noon, or I can set it to match at noon, and it is a poor fit for morning and afternoon. In practice I change the transmission value throughout the year, to give the best approximation. At the moment I'm using 0.86.

The variations due to clouds, know as Cloud Edge Effect in the world of PV solar, is a largely unrelated effect. Bright clouds due to reflection and forward scattering can greatly increase the radiation levels- it's like having a 2nd, weaker, sun in the sky. I've seen over 1700W/m^2 on my Davis VP2, and 1300-1400 is quite common when the sun is high in the sky.

Generally the theoretical curve is a very good fit to my home brew sensor, apart from as you say very low sun angles where it under estimates. It cuts off to zero at sun rise/set whereas my sensor can get a reading then particularly if the sky is cloudy or hazy. The model is not perfect and though it takes extinction into account does not seem to model scattering very well, and scattering when the sun is low is a significant component. Nor does it allow for scattering at all when the sun is below the horizon but still illuminating the atmosphere above you.

I have calibrated for the middle of the day where the model seems most accurate and let it over read morning and evening.