meteohub.de

I have a modifying factor of 0.342 on rain0

I can manually calculate the recorded rain over a period by checking raw sensor readings then adjusting by that factor.

If I then compare that to data output by Meteohub in various forms I find discrepances:

a) The logging protocol is producing the correct total rainfall figures in WD. :)

b) The WSWIN export is producing low figures. :(

c) [day1_rain0_total_mm] is producing the same figure as the (export based) WSWIN 'today'. :(

Taking a quick look at the WSWIN export lines, I notice that every tip now seems to be reported as 0.3mm, which would lead to a rounding error.

I'm sure that earlier exports used to produce a tip pattern in which there was the occasional 0.4mm.

Is this possibly related to the recent changes in rainfall handling?

As small worked example (which I meant to include earlier):

From midnight to a point this afternoon the (raw) sensor accumulated rain changed from 9120 to 9260, i.e. 14.0mm, which with the calibration factor becomes 4.8mm.

For the same time span:

WD (using logging protocol): 4.8mm

WSWIN (using export csvs): 4.2mm

[day1_rain0_total_mm]: 4.2mm

I observer that 4.8mm represents 14 tips, and that 14 x 0.3 gives 4.2mm, which ties in exactly with my 'rounding theory'...

your beta includes a new way of applying calibration to rain total. I will increase resolution to avoid the rounding issue with the next official update.

Thanks for pointing to this. It just effects computed data, so a recompute will fix this for your historical data.

admin wrote:your beta includes a new way of applying calibration to rain total.

Ah. Sorry. I forgot I was using a beta with respect to the rain calculations. I should have queried this privately.

...will increase resolution...next official update...recompute...fix..historical data.

Thank you for your prompt attention.

admin wrote:your beta includes a new way of applying calibration to rain total. I will increase resolution to avoid the rounding issue with the next official update.

Fine in the official 2.6b.

I see that you are now providing rainfall to 2 decimal places in the export files. That seems to work just fine for WSWIN export. Thank you. :)