Potential barometer issue (Meteostick)

[Brezentrager]

Hi,

AWEKAS doesn't accept my barometer figures since they differ too much from comparable stations in the region.

When looking at the issue in more detail I found a suspicious pattern for my barometer data - see screenshot of last week's pressure.

There's quite a big variation between the barometer figures in the morning (high) and in the afternoon (low). The variation is as big as 5 mbar - every day.

When comparing this to a station in the neighborhood, those variations are absolutely unusual.

When comparing the values to stations in the region, the lower afternoon values seem to be more comparable (=correct).

I know that the barometer values have some temperature correction applied to it. But why is the effect so huge?

The pressure and altimeter values look much more reasonable.

Did I set up my system in the wrong way?

As background:

The Meteostick is connected to a RPi (running the latest version of weewx and Matthew Wall's meteostimk driver).

Hints and tips appreciated!

Best,

Manfred 

[Pat]

Just curious how these values compare against your station's console?

[Brezentrager]

The values on the console are generally lower. Currently - Console: 1015.1 // WeeWX: 1017.9

The difference gets bigger in the morning and smaller in the afternoon/evening.

[Pat]

Actually this is probably a bad comparison since I think the Meteostick comes with it's own barometer on the stick itself - so the values won't be identical.

[Brezentrager]

Yes, the pressure values come from the Meteostick. I'm less concerned about the actual level of the barometer values - that's something I can calibrate.

I'm more concerned about the ups and downs that heavily correlate with my temperature and humidity values. 

[Pat]

My initial question was to try and isolate the readings (to see where the problem is)... Does the Meteostick have a way to see its values without weewx?  (web ui? serial port?) Trying to see if the stick is the problem or if it's weewx.

The driver looks like it expects "pressure" from the Meteostick, and submits that to the loop. As I understand it StdWxCalculate will calculate the barometer based on the pressure value in the LOOP. 

[Brezentrager]

Yes, that's also my understanding.

The "pressure" line in my screenshot should be the original pressure reading of the stick, isn't it?

My problem is that I don't know enough about what "magic" weewx applies when calculating the barometer values out of pressure.

When looking at the code of weewx (see below) I can see that the weewx calculation has quite an effect on the barometer value.

A temp-difference of 10C results in approx 2.5mbar difference in 'barometer pressure'.

My values therefore seem to be "correctly calculated". 

Still, I'm surprised to see this up and down. But when looking at the code, this obviously has to be that way. Whenever the temperature changes, the barometer value will also change considerably.

All other Weewx users should see the same pattern in their barometer data. Is that the case?

Manfred 

[Pat]

We may need Tom, mwall or Gary to offer their insight on the weewx internal code question.

I'm using the Meteostick driver as well however my hardware is a Moteino (Arduino) with Meteostick emulated code on it. So my hardware is not an exact like-for-like with your setup, however my driver is the same. I don't seem to be having the issues you are reporting. My values all seem to be within their respective boundaries? See my attachment which is in inHg and mbar for comparison. 

I think your question is valid - if this was a problem with weewx it would be a much bigger problem. Which goes back to seeing how you can troubleshoot your Meteostick directly. 

Every 30 seconds it outputs a "B packet" which contains the pressure of your station. We could modify your driver to log this to a text file so you can do a manual check of those values over the course of a few days? Not sure how else to troubleshoot this. 

[Brezentrager]

That's really weird. Your barometer line is absolutely parallel to the other two lines. 

But when looking at the weewx code, the effect of the temperature calibration on 'barometer' should definitely by bigger.

Ahh .. maybe that's the reason: Belchertown is only at a height of 180 meters! I'm at 555m

10C at your location only means +/- 0.8 mbar. At my location that's 2.5 mbar

The temp correction on my height seems to be quite extreme. The effect of the temp correction is much bigger (over one day) than the natural movements of the pressure. So my chart will always look like this - with all those daily ups and downs.

Would really be great to get some insights from the real weather experts here in the group!

As a short term fix I'll switch to 'altitude' on my webpage.

Thanks for your time!

[Michi Kaa]

Where is the Barometer located? Indoor? Since temperature and humidity influence measurement of pressure, it is possible that the measurement is just wrong just because you are opening a window for ventilating the house or when the barometer sensor is hit by direct sunlight during the day and is warmed up.

[Michi Kaa]

Am Freitag, 9. August 2019 07:35:46 UTC+2 schrieb Michi Kaa:

Where is the Barometer located? Indoor? Since temperature and humidity influence measurement of pressure, it is possible that the measurement is just wrong just because you are opening a window for ventilating the house or when the barometer sensor is hit by direct sunlight during the day and is warmed up.

The fact this happens repeatedly more or less the same time during the day lets me think its sunlight. How was weather on Aug 3 in Belchertown? Overcast?

[Brezentrager]

Good idea! It actually could be sunlight. The time (late afternoon) would fit. And yes, August 3 was a day without much sun. So was Aug 7, where the effect was also less pronounced.

Today will be a sunny day, so I'll move the Meteostick a bit into the shadow and see what happens.

[Andrew Milner]

I would make sure it is completely in the shade at all times.  I see the meteostick manual even says to ignore indoor temperature and humidity as humidity will always be 30% and temperature will be reading high due to temperature rise from the electronics!!  Goodness knows what warmth from the sun does to readings!!!!!!  Deffo keep it in the shade for any chance of sensible readings from onboard sensors I would think!!!!