meteohub.de

Hi,

I'm starting to find my way around WSWIN. Today I've noticed a possible discrepancy between the rainfall recorded in Meteohub and the rainfall passed on to WSWIN via the CSV export files.

The raw log for rain0 shows 3 rain events just after 08:21 UTC

20080629082133 rain0 0 0 5760
20080629082220 rain0 010 0 5770
20080629082307 rain0 010 0 5770
20080629082354 rain0 010 0 5780
20080629082528 rain0 010 0 5780
20080629082702 rain0 0310 0 5790
20080629082749 rain0 0310 0 5790

N.B. The raw rain totals need to be multiplied by a factor of 0.342

EXP06_08.csv (and presumably the import.csv that actually transferred the data) shows 4 rain events at that time (i.e. just after 09:21 BST)

29.06.2008,09:21,18.1,12.0,11.7,12.7,12.9,14.9,16.3,12.9,73.0,84.0,92.0,87.0,83.0,78.0,992.2,0.0,19.1,237,22.7
29.06.2008,09:22,18.1,11.5,11.4,12.7,12.6,14.9,16.3,12.9,73.0,82.0,91.0,87.0,83.0,78.0,992.2,0.3,22.0,252,36.7
29.06.2008,09:23,18.1,11.3,11.1,12.5,12.2,14.9,16.3,12.9,73.5,79.5,89.0,86.5,83.0,78.0,992.2,0.4,28.1,246,28.8
29.06.2008,09:24,18.2,11.2,10.9,12.4,12.2,14.9,16.3,12.9,74.0,79.0,88.0,86.0,83.0,78.0,992.2,0.4,24.3,267,28.8
29.06.2008,09:25,18.2,11.2,10.7,12.2,11.6,14.9,16.3,12.9,74.0,80.5,89.0,86.0,83.0,78.0,992.2,0.0,24.8,254,31.7
29.06.2008,09:26,18.2,11.0,10.6,12.1,11.5,14.9,16.3,12.9,74.0,83.0,90.0,86.0,83.0,78.0,992.2,0.0,22.0,249,20.5
29.06.2008,09:27,18.2,11.0,10.6,12.0,11.6,14.9,16.3,12.9,74.0,84.0,91.0,86.0,83.0,78.0,992.2,0.3,20.0,231,25.6
29.06.2008,09:28,18.2,11.0,10.7,12.0,11.8,14.9,16.3,12.9,74.0,85.5,92.0,87.0,83.0,78.0,992.2,0.0,22.7,260,24.1

The second rain event (i.e. 20080629082354) results in 0.4mm in the line for 09:23 and then apparently again in the line for 09:24.

<aside>
I notice that the the rain gauge did not report during the minute 08:24 UTC, but I can't tell if that might be relevant.
</aside>

All other outputs that I've checked (mainly web pages & HTTP logging protocol), report the rain as expected (which is how I came to spot this).

Any ideas?

PS. Using Meteohub v2.4e

Further to this...

I just did a daily comparison for the current month (June). The data exported by Meteohub to WSWIN includes approx. 10mm more rain for the month to date than Meteohub itself reports.

On most days the difference is perhaps a single 'tip' (i.e. ~0.3-0.4mm with my equipment), but on some days the difference is more significant, e.g. on the 24th, 8.5mm in Meteohub but 10.5mm exported to WSWIN.

The HTTP protocol 'exports' to WD agree with Meteohub.

If there is any data you'd like to take a look at to help sort this out Boris, you are welcome.

PS. The WSWIN 'File Watching' is really neat. I hadn't appreciated that it could automatically handle catch up of data logged while WSWIN was not running. :)

skyewright wrote:<aside>
I notice that the the rain gauge did not report during the minute 08:24 UTC, but I can't tell if that might be relevant.
</aside>

I think it probably is...

I noticed that some rain figures back in March were way out (e.g. 350mm on a day that had 2mm actual!).

Back in March I was having problems with rain gauge reception. The next 3 lines are consecutive actual sensor readings for rain0.

20080324121659 rain0 0 0 580
20080324124337 rain0 0 0 600
20080324153819 rain0 0 0 600

So 2mm if rain was noticed at 12:43, then the next rain0 reading wasn't received till 15:38.

EXP03_08.csv includes 2.0mm of rain for 12:43, and also for every minute from there to 15:37... :S

Perhaps that's something to checkout...

skyewright wrote:So 2mm of rain was noticed at 12:43, then the next rain0 reading wasn't received till 15:38.

EXP03_08.csv includes 2.0mm of rain for 12:43, and also for every minute from there to 15:37... :S

I don't wish to nag, but have you had a chance to look at this 'straight lining' of non-zero rainfall in the WSWIN export files?

Thanks for pointing out the problem in that detail. I will have a look at this the next days.

admin wrote:
I will have a look at this the next days.[/quote]
Thank you. :)

admin wrote:I will have a look at this the next days.

Still not nagging - just another observation to add. ;)

Something else I have noticed with WSWIN export.

If a sensor[1] is not available for a few hours, after a while the "on going" WSWIN export csv starts to have blank entries rather than 'straight line' figures for that sensor. This gives very strange graphs in WSWIN (e.g a jump to 0C and 0% humidity).

If after the sensor is functioning again the export CSVs are regenerated in Meteohub by doing a Save on the "Export of Weather Data in WSWIN Format" page, the fresh CSVs 'straight line' right though the time when the sensor was not available - which is what I'd expect, and prefer the 'on-going' data to do too[2].

[1] I've only actually seen this with TH and T sensors.

[2] Except of course for the rain sensor! :laugh:

I did some changes to how data is printed out the WD and WSWIN with the just released 2.6a update. I hope this will improve on the situation. I have to admit, that I didn't had much testing on this. If you like, give it a try.

admin wrote:If you like, give it a try.

Will do. I'm keen to have the WSWIN feed going well, so that I can consider making it my main weather data analysis application.

admin wrote:If you like, give it a try.

Similar behaviour still observed. :(

Here's the new equivalent of the June 29th export I quoted earlier:

29.06.2008,09:21,18.1,12.0,11.7,12.7,12.9,14.9,16.3,12.9,73.0,84.0,92.0,87.0,83.0,78.0,992.2,0.0,19.1,237,22.7
29.06.2008,09:22,18.1,11.5,11.4,12.7,12.6,14.9,16.3,12.9,73.0,82.0,91.0,87.0,83.0,78.0,992.2,0.3,22.0,252,36.7
29.06.2008,09:23,18.1,11.3,11.1,12.5,12.2,14.9,16.3,12.9,73.5,79.5,89.0,86.5,83.0,78.0,992.2,0.3,28.1,246,28.8
29.06.2008,09:24,18.2,11.2,10.9,12.4,12.2,14.9,16.3,12.9,74.0,79.0,88.0,86.0,83.0,78.0,992.2,0.3,24.3,267,28.8
29.06.2008,09:25,18.2,11.2,10.7,12.2,11.6,14.9,16.3,12.9,74.0,80.5,89.0,86.0,83.0,78.0,992.2,0.0,24.8,254,31.7
29.06.2008,09:26,18.2,11.0,10.6,12.1,11.5,14.9,16.3,12.9,74.0,83.0,90.0,86.0,83.0,78.0,992.2,0.0,22.0,249,20.5
29.06.2008,09:27,18.2,11.0,10.6,12.0,11.6,14.9,16.3,12.9,74.0,84.0,91.0,86.0,83.0,78.0,992.2,0.3,20.0,231,25.6
29.06.2008,09:28,18.2,11.0,10.7,12.0,11.8,14.9,16.3,12.9,74.0,85.5,92.0,87.0,83.0,78.0,992.2,0.0,22.7,260,24.1

The rain0 tip in minute 23 is again duplicated in minute 24 (in which there was no sensor reading from rain0)...

ok, I hope I have finally detected the reason. You have mail...

admin wrote:ok, I hope I have finally detected the reason.

I believe you have. As a result of your e-mail rain exports are now looking good. :cheer:

Thank you very much for your efforts.