Main Page Descriptions

[Super-T]

Steve
I was comparing the readings on the wh1081 console and noticed that the "Wind" on the console was showing quite a bit less than was showing in Cumulus as "Latest". The booklet for the wh1081 describes "wind" as the average Wind.
Allowing for different update times between the console and Cumulus, it is hard to figure out what the actual windspeed is at the time.
Does the Cumulus "Latest" mean, "what the actual windspeed on the console was showing"
Basically the question is, "what does latest mean in Cumulus"?
Is there a list somewhere describing the terms used on the Cumulus main page?

[steve]

There isn't a list as such, but I've discussed this in various places in the forum. I really should write some documentation which describes how the various parameters are derived, but it's finding the time. And I'd still get people asking for the information anyway

Here is what I said recently about this:

Cumulus reads two values from the station every 15 seconds (this figure is for Fine Offset stations, though they might not get updated in the station that often, I think it may be about 48 seconds). Let's call them 'station gust' and 'station speed '. Exactly what these values really represent depends on the type of station, and I don't know how Fine Offset stations calculate them.

Cumulus displays each 'station gust' value as 'Latest'. It displays the highest of these from the last 10 minutes as 'Gust'. If you don't have the 'calculate 10-minute average' option set, Cumulus displays the 'station speed' figure as 'Average'. If you do have the 10-minute average option set, it calculates a running 10-minute average of the 'station gust' figures and displays that as 'Average', and completely ignores the 'station speed' figure.

Of course, whether Cumulus is doing sensible things with the two values that the station supplies depends entirely on how they are derived. Dave Cornwell recently posted here a response he got from Fine Offset about this, but it was rather inconclusive. It looks to me from their response that perhaps the 'station speed' value is an average over the last 48 seconds. But what is 'station gust'? Does looking at the console and comparing 'gust' and 'wind' (is that how they are labelled?) give any clues?

Steve

[Super-T]

The manual I have describes the wind reading on the console as average and the comms from the transmitter as 48 seconds.

I'll play with the 10 minute average thanks

[steve]

Bear in mind that when reading the data from the logger at startup, there won't be many readings per 10 minutes (max 2) to create a 10-minute average from.

[dc1500]

I'm getting a bit of a headache with this as well now! After the prompt reply to my first email I sent Fine Offset another which basically asked for a slight clarification of what they said plus "how do you calculate the gust then?" This time no reply so far! I sat and stared at the console and decided the windspeed changed every 48 secs and so did the gust value. My theory at the moment is that it measures the number of revs of the anenometer for 48 secs (why not 60 secs?) and then converts that to a wind speed. So if 100 revs in 48secs = 5mph, say, then 200 revs in 48 sec = 10mph and so on. They then said "so it is relatively small (48 secs) comparing to instant wind speed or averaged wind speed in a shorter time. "Then maybe a gust value is calculated by the highest of a shorter measuring span of say 10 lots of 4.8 secs and changed every 48 secs. The display can be toggled between either gust or average. (I'm not sure either how that will affect the capture to Cumulus) All I can say is that the gust is about 1 to 2x times higher than the average. Incidentally, do the units have to be the same in the console as in Cumulus or does Cumulus read the units and then convert? I think I will have to sit down and write things down and try and work it out scientifically. I'll post on here if I get a reply. It's mainly of academic interest only because the siting of mine is probably giving rubbish results in terms of accuracy anyway.
Dave C

[steve]

Yes, this is the problem I have in trying to guess what their 'gust' could be, if their speed is an average over 48 seconds (no doubt calculated in exactly the way you suggest). As you say, you might imagine that their 'gust' could be the highest value of some set of smaller periods, but this would require more measurements to be taken, and they haven't mentioned anything about this.

It would be good to get definitive answers about this, because it may need a re-think on how I use the values in Cumulus. I'm already feeling a little uneasy about it.

Steve

[dc1500]

If you don't do anything to the raw data being sent can Cumulus distinguish between Fine Offset's version of a gust and their version of wind speed? In other words the two values we toggle between. If so, it might be best to just leave it to us to fiddle with calibration settings, at least till we found out how the gust is calculated.
Dave

[steve]

I'm not sure what you mean by the first part.

As long as you don't get Cumulus to calculate a 10-minute average, there's not really an issue. Cumulus will display the 'gust' value as 'latest' and display the highest of those in the last 10-minutes as 'gust'. It will display the 'speed' or whatever we're calling it - the 48-second average thing - as 'average'. That all seems reasonable to me.

The issue comes if you have 'calculate 10-minute average' set, because Cumulus uses the 'gust' values to calculate the 10-minute average. On other stations, this is a reasonable thing to do, because the 'gust' value on those is really just the speed calculated over the last few seconds. It seems to me that for Fine Offset stations, it should really be using the 'gust' value just for gusts, and the 48-second average to calculate the 10-minute average. But it would be nice to know what the 'gust' value means before I change anything, and as I keep saying, there are issues with getting Cumulus to calculate a 10-minute average anyway when the data is from the logger, when the data samples are 5 minutes apart at best.

Steve

[Super-T]

I suppose it is possible that the transmitter looks at the highest wind speed in the 48 seconds and then transmits it? If it does then the console would have no need to calculate the gust.
Gusts at my place seem to last less than 48 seconds so it is possible the transmitter stores the info for transmitting.
A viewport into the transmitted data would be handy, along with a description
Hopefully Fine Offset are watching this forum!

[dc1500]

Sorry Steve, I meant in the first part if Cumulus doesn't calculate the average. The rest you have clarified and I agree, it's not an issue, well only an academic one to find out what Fine Offset do in their calculations.
Dave C

[steve]

I suppose it is possible that the transmitter looks at the highest wind speed in the 48 seconds and then transmits it? If it does then the console would have no need to calculate the gust.

Yes - but it would have to have to have some sort of mechanism for measuring it. All these things can do is count rotations over a given period of time, and all we know so far is that it appears to count the rotations over a 48 second period and transmit that as the average speed. It must be actually counting rotations over smaller periods (as Dave suggested), and as you say, sending the highest of those every 48 seconds.

Steve

[steve]

Dave - I realised that I didn't answer your question about units. The station supplies the data in fixed units, regardless of what console settings you have, so Cumulus just converts from those units to whatever units you have configured in Cumulus. And then it stores the data in those units in the data logs, which is the stupid part

Steve