monthly file

[serowe]

Steve - wasn't sure whether to post this in Cumulus or Cumulus 2 - so, her 'tis.

I know that Cumulus 'centre's' the wind direction at N or S (depending on the config) for calm conditions, but is it possible that this direction not be exported in the detailed monthly log as 'N' or 'S'? I can handle it programmatically if needed, but after processing recent records I suddenly realised no, we weren't getting all those northerlies (which are usually warm in winter, hot in summer here) as the conditions were calm. As I said - I can handle it, but before I start doing something like that I just want to check what is also in the pipeline for V2 in case I get caught.

Also - the daily file doesn't export a number of fields such as max and min humidity, dewpoint, indoor temp and humidity, wind chill, heat index - will these be included in any future exports? There is obviously a record of indoor details and humidity (as they are graphed) as are other readings which, in some parts of the world, we never see (air frosts and snow for example - uinless you want to box some up and post it over )

I only noticed a number of these as I now have some scripts running which process the MySQL database of all my past data, but I noticed when I got to data that came from Cumulus there were a number of columns of data missing, and then realised why...

No drama, just checking so I work out what to do with the scripts I've written.

http://mtb-images.dyndns.org:1088/weather/history will give you an idea of what I am currently doing with the historical data. First page is from dayfile - just a daily summary by month - each date can then be clicked on to bring up the detailed records for a single day. I'm still working on an Annual Page (which will, eventually, become the first page so you go straight into year, then select a month, then select the day. Nowhere near finished, but these two pages are currently working reasonably well.

[steve]

The history of Cumulus is basically that I decided to pretty much stop development of Cumulus 1 about a year ago, as it clearly needed several facilities that weren't in there originally, nor planned for, and it was getting very hard to do major changes and keep it relatively stable. So this is why it is lacking lots of things that you might think were pretty basic.

I bravely/foolishly started on the complete rewrite that is Cumulus 2 so that I could then go on to add the missing facilities. It's just taking a long time...

Cumulus 2 addresses most/all of the items you mention, or should do when it reaches its first proper release.

[serowe]

Steve - no problem - don't get me wrong, not complaining, just asking so I know where to head with my own scripts. Personally I believe in 'less is better' in most cases - less to go wrong when you do change something

[steve]

Steve - no problem - don't get me wrong, not complaining, just asking so I know where to head with my own scripts.

I understand completely. I am not as defensive or paranoid as some people think - Objective, constructive criticism (in the the correct sense of the word) is good; I "tell it like it is" and I am happy for Cumulus users to do the same.

[steve]

I know that Cumulus 'centre's' the wind direction at N or S (depending on the config) for calm conditions, but is it possible that this direction not be exported in the detailed monthly log as 'N' or 'S'? I can handle it programmatically if needed, but after processing recent records I suddenly realised no, we weren't getting all those northerlies (which are usually warm in winter, hot in summer here) as the conditions were calm. As I said - I can handle it, but before I start doing something like that I just want to check what is also in the pipeline for V2 in case I get caught.

This has been going around in my head today, and I've realised that I don't actually understand what it is you're saying. I thought you were talking about the graphs, where 'calm' gets plotted incorrectly as 'North' (but never South), but it looks like you're actually talking about the data logs? But Cumulus doesn't put 'N' or 'S' in the logs? It puts in a bearing in degrees - zero for calm, 180 for South, and 360 for North. This is the standard according to the UK Met Office, and also, I think, the WMO.

[serowe]

I can't recall off the top of my head what the BoM in Oz use (I worked for them for nearly 20 years!). But checking my old data files from VWS, WD and FreeWx, they all use the last bearing the wind came from.

The scenario I am seeing in the data is like this: I have a very light breeze coming in from say WSW. For a couple of recording points there is nothing - but the data suddenly shows 0 (zero) and, if I try to analyse it, it appears that the wind vain has suddenly swung around to the north. Now I kow that the wind speed reading is also zero BUT I have instances where there is a gust within the one minute (or whatever the time period the log has been recording at) of some positive amount - again, inthis example, say from the WSW, but still the recorded data shows it is still coming from the N.

I guess what I (personally) would prefer to see is, sure, maybe centre the wind gauges at one cardinal point, but the data shoud reflect either a) the last point it came from or b)nothing at all - not zero OR a value that indicates nothing but cannot be taken as a cardinal point - does this make sense now?

However - writng this even the gauges should, IMHO, not be changed to show directions. Where I live physically, can be quite sheltered. For instance, today we have a weather warning out for winds to 65 km/h, very heavy rain possible, flash flooding etc. I will be lucky (well depends on how you look at it) to see the winds rise above 25-30 km/h. But I will also see many periods of no wind - the wind is from an intense low almost over us so we can expect wind from the east, south and through to the west as it comes through BUT if it is calm, the data will show from the north (and the news has just reported a tornado on Victoria's northern border about an hour ago - oh boy, wild day ahead!) In fact, just checking the Official records they use the word 'Calm' when there is no wind - a much better solution, again IMHO .

Finally - and again, this is a personal POV, by placing a zero in the data, ity is effectively recording false data.

[steve]

I agree that gauges should probably stay pointing to the last position they were pointing to (like a real wind vane), and to be honest, I can't remember what the various gauges in Cumulus and on the web actually do! I suspect they probably do interpret zero as North (like the graphs) and hence move to the top, which isn't ideal.

But the data logs are correct (aside from any bugs in the implementation). Zero, according to the UKMO and (I think) the WMO, means 'calm', not North. 360 means North. If there are countries where zero doesn't mean calm, but North, then I can put this an option in Cumulus 2.

[serowe]

If this is the way you have interpreted it (ie 0 as calm and 360 as North, I can work with that. I can't remember what we used to code back in the 80's (I just remember having to key hundreds of 5 digit groups!).

I can easily modify my code for the Cumulus 'standard' and will be doing that this morning - the earlier data I have will possibly be a problem, but I can work around it

Maybe, for V2, the option alongside deciding to point the dial to N or S could include a 3rd option 'Do not adjust', then that would cover all bases.

[serowe]

This is how I have ended up handling it - times from 0000 to 0840 were calm, the entry at 0841 shows a gust, and then back to calm again:

http://mtb-images.dyndns.org:1088/weath ... =08&day=10

[steve]

If this is the way you have interpreted it (ie 0 as calm and 360 as North,

It's the way the UKMO specify. Generally, I try to stick to standards. I got annoyed with other software which allowed pretty much anything whether it made sense or not, which was one reason I wrote my own software. I'm not a meteorologist so I sometimes don't know what is the correct thing to do, so I follow UKMO standards as much as possible.

the Cumulus 'standard'

That's flattering, but as I keep saying, it's the standard of the UKMO.

[steve]

This is how I have ended up handling it - times from 0000 to 0840 were calm, the entry at 0841 shows a gust, and then back to calm again:

Yes, that's the correct thing to do. Zero is a special value which means calm, and you have interpreted it as such. You would have to do exactly the same thing whatever the special value was.