Cumulus Support

Hi There.

Had a spike in my readings overnight yet cumulus still recorded this even though I had the spike setting to 15C


Heavyweather stats as follows:

7/3/12 2.00am 16.7c
7/3/12 2.20am 16.5c
7/3/12 2.40am -7.0c
7/3/12 3.00am 15.9c
7/3/12 3.20am 15.8c

My understanding was the spike reading would not be counted if it fell outside my spike range (i.e 15c)

Given it dropped 23.5c, why has this not worked ??

I presume it has nothing to do with the positive/negative temperate and somehow assumed a drop of only 9.5c (16.5 - 7.0)


Any help appreciated.

The most likely explanation would be that it didn't change 'suddenly', i.e. in one step. If you zip up the diags folder and attach it, I'll see if I can see anything.

I do sometimes get spikes, but the spike removal does work with when you download from the logger as I found out the hard way here: https://cumulus.hosiene.co.uk/viewtopic.php?f=15&t=6846 . Did you download from the logger, you gave the figures from the logger (heavyweather history file). I understand it only works on live data.

Matt.j5b wrote:I understand it only works on live data.

This is true, as I've mentioned many times in the past in the forum. But the help doesn't make this clear; I'll correct that.

One of the reasons that spike removal doesn't operate on logger data is that it's not practical to apply the same spike removal values to data that is several minutes apart (up to an hour for some people). Historically, the spike removal was for Fine Offset stations only, and data from the logger doesn't seem to suffer from the same corruption that live data does.

steve wrote:... and data from the logger doesn't seem to suffer from the same corruption that live data does.

Well if it is only logging 1 in 25 readings say for a 20 minute log interval, the chances of logging a duff value are reduced by the same factor.

steve wrote:data from the logger doesn't seem to suffer from the same corruption that live data does.

I run off the logger at 5 minute interval a lot of the time and don't get many spikes because there is less chance of it happening rather than when its running live. I am happy with that, and what you say makes sense.

I just wondering whether dids04 got it from the logger because logger figures from heavy weather were quoted in the first post.

Thanks for the responses all

Steve, I have attached a copy of the Diags folder but as you suggested, data was obtained from the logger.

I don't run live but rather on a 20 minute setting, which I would have thought would make spikes quite rare.

This is the 2nd low temp spike I've had since setting up a few weeks ago.


Thanks

dids04 wrote:data was obtained from the logger.

Yes, so as explained, the spike removal facility doesn't apply. Maybe one day Cumulus will have more sophisticated erroneous data removal that also works for logger data.

steve wrote:

dids04 wrote:data was obtained from the logger.

Yes, so as explained, the spike removal facility doesn't apply. Maybe one day Cumulus will have more sophisticated erroneous data removal that also works for logger data.

I hope so

Some days ago, while I was in a job travel my Weather Desktop hang out for several days, When I return to home and reactivated the PC, Cumulus effectively downloads the data logger data, but plagued with the well-know spikes, ruining my data ...uggg... By the way, my data logger is preset for 1 minute readings...