Solar Settings

[lairgweather]

Has anyone discovered the best solar settings to use to actually record when the sun is shining?
Also is there a webtag for Twitter to post todays sun hrs
thanks

[steve]

Part of the problem is that the ideal settings will vary depending on your location. I'm getting quite good results with

SunThreshold=75
RStransfactor=0.86
SolarMinimum=9

This is with a Davis VP2.

[lairgweather]

Cheers Steve i will give that a whirl.
Is there a webtag for Twitter that will post 'Todays Sun Hrs?'

[steve]

<#SunshineHours> - http://wiki.sandaysoft.com/a/Webtags#Da ... Sun.2FMoon

[Hillbilly]

Part of the problem is that the ideal settings will vary depending on your location. I'm getting quite good results with

SunThreshold=75
RStransfactor=0.86
SolarMinimum=9

This is with a Davis VP2.

Thanks Steve, I asked the same question in my slightly later post which crossed with this. Do you have a view on what degree of difference I may need to use on these settings in NW France compared to you? I presumed location plays a part but I have no idea whether there is a difference between Scotland and Southern England, or whether Northern Europe would all be similar.

[steve]

It's probably as much a mystery to me as it is to you. I got the algorithms for calculating theoretical max solar radiation from the spreadsheet here: http://www.ecy.wa.gov/programs/eap/models.html - the 'solrad' link. I used the Ryan-Stolzenbach formula. The spreadsheet suggests values between 0.7 and 0.91 for transmission factor, with a default of 0.8. I tweaked that value as best as I could until my maximum actual readings compared reasonably well with the theoretical calculated max on a bright sunny day, trying to make it never exceed the maximum. I then played with the other values until the sun on/off detection seemed to correspond to what the sun was actually doing.

[Hillbilly]

Ok, thanks Steve. Just equipping my senior business user with a pad, pencil and the spreadsheet and sending him off to make observations!

[colinpb]

Firstly how accurate do you want your recordings to be. If a general impression is all you are after, then altering with the settings will be fine. To be as accurate as the 3080 will allow will take time.

I decide on as accurate as I could get, and I've been fine tuning my 3080 since mid-November 2011. After reading as much as I could find I come to a number of conclusions.

Firstly what sunshine did I wish to record? Statistically the UK MET Office records Bright sunshine; this enables comparison with historical records. These measurements are made using a Campbell Stokes recorder: 19th century technology. However it cannot record all sunshine. I decided that in the 21st century I wanted to record both Bright sunshine (BS) and Total sunshine (TS).

The sensor on the 3080 is extreme lateral thinking, it's a lux meter not a solar sensor as on a Davis station; the Davis sensor costs as much as two complete 3080s. However I felt there was meaningful data available from the 3080 if I put in the time.

I realised attempting to get the greatest accuracy by constantly changing cumulus settings would be like stabbing in the dark. I would end up with data sets based on various settings, which would be impossible to make any sense of.

Finally I have a faulty sensor, and there’s a good chance the replacement would mean another two months repeating my work to get data from the old and new to align.

So I decided on the following path.

I take copies of the monthly log files and convert to excel. I then apply formulas to the solar readings that allow me to mimic the cumulus sunshine records. I have one column for BS (bright sunshine) and another for TS (total sunshine). I can alter values for threshold, limit and conversion factor and this changes values over the complete log file. It then enables me to see if my observations for sun/cloud are matched by a corresponding value at the same time period in the spreadsheet. I also noticed that the solar readings appear to be snapshots, not rolling averages. Therefore I changed logging to 5 minute intervals to increase the chance of recording sun/cloud.

I have now decided cumulus will keep BS readings and I will keep a separate file for TS readings. The results from both will be manually entered into revised versions of the NOAA reports.

The settings I use for Bright sunshine are…. threshold 70, minimum 35, transmission 0.8. However I have also changed the conversion factor of .0079 to .0065 in the cumulus.ini as I was getting readings way too high. Based on the default .0079 conversion factor, my original Bright sunshine settings were ...threshold 85, minimum 35, transmission 0.8

Regards

Colin

[Hillbilly]

Hi Colin thanks for this detailed analysis. Like you we want to record data against fixed parameters rather than constantly moving goalposts. And the 3081 will only be good to a certain point. In the end, if our readings are consistently inconsistent that may be our best. Interesting point about the conversion rate we will look at that too.

Had a downpour this morning and lost solar readings for an hour or so afterwards. Deep joy!

Will post results as we experiment.

[Flosex]

Thanks for your interesting and helpful posting, Colin.

I have just started using a WH3080 and have set it up to record sunshine as per the 120/m2 criteria. However, I have noticed that when the sun is shining quite brightly, the m2 reading is quite often under 120, and Cumulus is not recording this as sunshine.

I think I would therefore like to record total sunshine (TS), and wondered what your settings are to record this, please?

Thanks and regards.

Chris

[steve]

I have just started using a WH3080 and have set it up to record sunshine as per the 120/m2 criteria. However, I have noticed that when the sun is shining quite brightly, the m2 reading is quite often under 120, and Cumulus is not recording this as sunshine.

You can't use the standard fixed 120 W/m2 value for determining sunshine, because that figure refers to direct radiation, and the sensor doesn't measure that, it measures global radiation. This is why Cumulus has all of this complicated stuff based on your location and the time of day, and the day of the year.

This applies to the Davis sensor, but I assume it also applies to the Fine Offset one.

Note also that the Fine Offset sensor doesn't measure W/m2, as Colin mentions above; it measures Lux. Cumulus applies a simple multiplier to get a W/m2 figure from the Lux value, so there's another variable to consider. The default multiplier is 0.0079, which is also the figure suggested by Fine Offset.

[AllyCat]

Hi Chris,

when the sun is shining quite brightly, the m2 reading is quite often under 120,

Also, is the sunlight actually striking the solar detector (the little white dome in the Solar Pod) at the time?

Cheers, Alan.

[Flosex]

Thanks Steve. The display console of the WH3080 I have just bought (from Nevada this week) has an option to display Light as fc, W/m2 or lux. I assume from what you say, that the measurement taken at the sensor is in lux and the display converts it to W/m2 or fc if either of those options are selected?

We currently have sunshine in a clear blue sky and my display is showing 10.20k lux, 948 fc or 14.9 W/m2. I don't know about the fc reading, but the 14.9 W/m2 is low compared with the Cumulus reading of 10108 lux and 83 W/m2 (it has just gone a bit cloudy!).

Chris

[steve]

Thanks Steve. The display console of the WH3080 I have just bought (from Nevada this week) has an option to display Light as fc, W/m2 or lux. I assume from what you say, that the measurement taken at the sensor is in lux and the display converts it to W/m2 or fc if either of those options are selected?

Yes, that's right.

We currently have sunshine in a clear blue sky and my display is showing 10.20k lux, 948 fc or 14.9 W/m2. I don't know about the fc reading, but the 14.9 W/m2 is low compared with the Cumulus reading of 10108 lux and 83 W/m2 (it has just gone a bit cloudy!).

It looks like the console is using a much lower conversion factor than the 0.0079 that Cumulus uses. That's odd, because I know I have seen it documented somewhere that Fine Offset say that the value of 0.0079 is the one to use. I just can't remember where I saw that.

[AllyCat]

Hi Chris,

Yes, the Solar Pod transmits a "Lux" value, which as Steve said cannot be accurately converted to W/m2 for (at least) two reasons. Firstly Lux is a measure of visible light, but the W/m2 normally includes around 50% of Infra Red. Secondly, the FO sensor uses a diffuser, whilst the W/m2 value normally employs a (horizontal) flat-plate sensor.

But basically, your sensor seems to be under-reading considerably (more than the +/-30% specification quoted by FO). The Lux value in direct sunlight should be nearer 100k; clear sky in a shaded location is nearer 10k. Incidentally, what UV value is it reading?

Cheers, Alan.