Author Topic: Wind Gust variation between Cumulus and Ecowitt solved (Read 1796 times)

Waimarie · « **on:** March 14, 2020, 03:30:40 AM »

The output of my GW1000 goes direct to Ecowitt.net and also to Cumulus MX on a PC.
Why are wind gusts at least double the speed when shown on CumulusMX?
I have included 2 plots over the same day. Max gust on Ecowitt 11.2kph. at 11.05am Max shown by Cumulus 26.9kph at 2.50pm
Any idea why? and which is the closest to actual?

[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]

mcrossley · « **Reply #1 on:** March 14, 2020, 05:50:03 AM »

According to the Ecowitt documentation the GW1000 reports wind speeds in m/s, so both software are having to convert it to km/h. I just double checked my code and it looks correct. Of course if could be bug in the GW1000 as well - it is using two different protocols to provide the two sets of information.

What I'd need to see is the Cumulus MXdiags log file - with debug and data logging switched on for a period of time (say half an hour) - and the recorded wind speeds in Cumulus and Ecowitt.net for the same period.

What would also be useful as a sanity check are the wind speeds from your WS View app (do they match the live Cumulus values?) and maybe a nearby station if both yours and that comparison station have reasonably well mounted anemometers.

Thanks

Mandrake · « **Reply #2 on:** March 14, 2020, 08:42:20 AM »

Hi Folks, this one intrigued me so I have double checked and it looks OK to me.
I did some of the original alpha/beta testing for MX to work with the GW1000 so I had previously looked closely at this sort of thing.
See Ecowitt.net data for today and then MX data graphs.
[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]

I am fairly content that my data aligns on both, so it makes me suspect that you have a setting wrong somewhere!

Waimarie · « **Reply #3 on:** March 14, 2020, 02:38:26 PM »

Quote from: Mandrake on March 14, 2020, 08:42:20 AM

I am fairly content that my data aligns on both, so it makes me suspect that you have a setting wrong somewhere!

I was wondering about settings, must check it out more fully

Waimarie · « **Reply #4 on:** March 14, 2020, 02:43:25 PM »

Quote from: mcrossley on March 14, 2020, 05:50:03 AM

According to the Ecowitt documentation the GW1000 reports wind speeds in m/s, so both software are having to convert it to km/h. I just double checked my code and it looks correct. Of course if could be bug in the GW1000 as well - it is using two different protocols to provide the two sets of information.

What I'd need to see is the Cumulus MXdiags log file - with debug and data logging switched on for a period of time (say half an hour) - and the recorded wind speeds in Cumulus and Ecowitt.net for the same period.

What would also be useful as a sanity check are the wind speeds from your WS View app (do they match the live Cumulus values?) and maybe a nearby station if both yours and that comparison station have reasonably well mounted anemometers.

Thanks

I will get back to you on this. As luck would have it, the wind is dead calm at the moment.

Waimarie · « **Reply #5 on:** March 14, 2020, 07:45:50 PM »

AARRGGHH! Sorry folks. I missed the calibration section in Cumulus. Somehow multipiers had appeared in the wind section. I don't recall ever looking in that section before, but who knows? Anyway, long and short, it is now slightly breezy so I can re-check and both graphs now coincide. Stupid, but simple to fix.

the beteljuice · « **Reply #6 on:** March 14, 2020, 11:30:18 PM »

Quote

Stupid, but simple to fix

Except you're going to have to edit all your history !

Waimarie · « **Reply #7 on:** March 15, 2020, 12:49:36 AM »

Quote from: the beteljuice on March 14, 2020, 11:30:18 PM

Quote
Stupid, but simple to fix
Except you're going to have to edit all your history !

Shhhh. We'll just quietly sweep it under the carpet and forget it ever happened

News: