Sunshine hours in minutes?

[NigelG]

Is it possible to extract sunshine hours in minutes from Cumulux MX somehow? The reason behind this apparently contradictory question is my use of the somewhat idiosyncratic WsWin software for long term weather analysis. This requires sunshine duration to be provided from Cumulus MX to WsWin as the number of minutes recorded in each logging interval. Cumulux MX only provides a resolution of 0.1 hours in the standard monthly log. I am currently using the standard logging interval of 10 minutes. Changing my logging interval to a multiple of 6 minutes may provide a solution but I don't really want to mess with a good working system.

[water01]

You could use the webtag sunshinehours and multiply by 60.

[HansR]

Well, David is just ahead of me and I think that is the only solution.

Nevertheless, if you really want the sunshine per interval in minutes you will have to do some arithmetic as the webtag #SunshineHours only gives hours so far, you will have to make the difference with the previous value and convert that from hrs to minutes.

That works if you want to use webtags. But from your post I understand you want to use the monthly log. In there you find the value of the webtag in field 24 (hrs of sunshine today so far) and you will have to do the same calculation as with the webtag. So it depends a bit on how you are calculating.

There is a problem however, when you start analyzing the monthly logfiles you will notice that the interval in longer series is not constant and apart from not being constant you will find there are gaps. That will lead to errors in hrs of sunshine. I assume (not sure, neither do I know how the webtag values vary with disturbances in the interval value).

In my CumulusUtils I took another approach (because I also needed to calculate the energy for the whole day it was not really additional work):

1. Loop over all entries of a day and if the theoretical sunshine max > 0 use the row for calculation
2. Detect the interval size by calculating the time difference to the next row
3. Calculate if the sunshine is counted in that interval i.e. radiation > theoretical max radiation * SunThreshold (SunThreshold you find in settings)
4. If sunshine is counted, assume it is valid for the detected interval period so set your SunshineHrs = SunshineHrs + IntervalDuration
5. Goto the next interval

This will lead to a different figure than the number you find in the dayfile but it will be close. Errors will statistically even out. You will find that catch-up and start/stop isssues in the past (current CMX versions have this issue much less as it is much improved) may lead to gaps. If gaps are longer than 30 minutes I use to skip that period. Bad luck but that is life. No dat no value.

Well, many roads to Rome. Hope it helps a bit

[NigelG]

Thank you for your suggestions. I shall have to experiment with the theoretical and measured values on every sample to see how the result compares with the value derived by Cumulus MX. Hopefully, I can align the results on a sufficiently large sample.