Pressures Problems

[Drew S.]

Hi there, new user here.

I'm having trouble with pressures (what else, right? :) ).

I use a LaCrosse C86234, with what I think is the latest weewx (I don't know where to go to check the version).

My pressure problem is the pressures being reported to CWOP and WU are too high. The pressure being reported to WU is currently 34.95 inHg, while a nearby station is reporting 30.03 inHg. CWOP is receiving pressure as 1176.1 mb, whereas a nearby station is reporting 1018.0 mb. I have weewx posting to Twitter as well, using the [pressure] key, and it reports 29.83 inHg.

So... is this a problem with weewx's calculations, a problem with how it's set up, or a problem with my station? How do I fix it?

Links for comparison:
Mine:
https://www.wunderground.com/personal-weather-station/dashboard?ID=KCOCHEYE5
http://www.findu.com/cgi-bin/wxpage.cgi?call=DW8525

Nearby:
https://www.wunderground.com/personal-weather-station/dashboard?ID=KCOCHEYE2
http://www.findu.com/cgi-bin/wxnear.cgi?call=DW8525

[Thomas Keffer]

What does the console on the LaCrosse say?

-tk

Fat-fingered from my Android

--
You received this message because you are subscribed to the Google Groups "weewx-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to weewx-user+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Drew S.]

LCD says 29.79

[Thomas Keffer]

Ok. That is not too different from the nearby station. Check your altitude setting in weewx.conf.

-tk

Fat-fingered from my Android

[Drew S.]

Already have. I opened the GPS on my phone to double check. It is correct, 4250 feet.

You received this message because you are subscribed to a topic in the Google Groups "weewx-user" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/weewx-user/hFp03F9XVH4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to weewx-user+...@googlegroups.com.

[Thomas Keffer]

Please run wee_debug and send the results.

-tk

Fat-fingered from my Android

[Drew S.]

Here you go.

[Thomas Keffer]

Thanks.

A gauge pressure of 29.79 at an altitude of 4,200 ft results in an "altimeter pressure" of about 34.5 inHg, which is roughly what you're reporting to WU. That's very high, so I suspect the gauge pressure is wrong, or it's actually altitude-corrected pressure.

Does the LaCrosse console offer the opportunity to enter an altitude? Is it correct? If that's all OK, then we have a driver problem.

Matthew: just to check: the ws28xx driver returns gauge pressure, right?

-tk

[Drew S.]

No, there's no place to enter altitude into the console. But there is this in the manual:

[Thomas Keffer]

I see.  Did you do as they suggest? 

-tk

[mwall]

On Sunday, May 15, 2016 at 2:32:48 PM UTC-4, Tom Keffer wrote:

Does the LaCrosse console offer the opportunity to enter an altitude? Is it correct? If that's all OK, then we have a driver problem.

Matthew: just to check: the ws28xx driver returns gauge pressure, right?

unknown.  the hardware reports a 'PressureRelative' (that is the label used in the decompiled code), and the ws28xx driver uses that as 'pressure'.

although it is possible that the value is actually an adjusted 'barometer' value, the lack of any altitude in the station configuration makes it unlikely.

m

[Drew S.]

Yes. I used the value shown on WU from the other station in town. If you look at the graph at the bottom of https://www.wunderground.com/personal-weather-station/dashboard?ID=KCOCHEYE5 you can see when I did this.

[Thomas Keffer]

That's the conclusion I'm coming to.

Yet, they encourage the user to adjust the displayed reading to something closer to neighboring stations.

So, we need to either tell users "don't do that," or have the driver label the value as 'altimeter'.

-tk

--

[Thomas Keffer]

Thanks, Drew. I think we've solved our mystery.

The driver is expecting raw "gauge" pressure. That is, the actual, uncorrected pressure at your station. But, the user's manual encourages users to adjust the value to reflect altitude. Unfortunately, it's just a simple offset, not a true altitude adjustment. So, what's coming out of the driver is pseudo altitude-corrected value. The driver doesn't know this, and adds another altitude (this time, a true!) correction, resulting in abnormally high pressures.

As I noted to Matthew, we have two choices:

1. Do not adjust the value. Leave it at the raw gauge pressure.
2. Change the driver to have it think it's getting an altitude-adjusted value. Because the instrument is not really emitting a true, altitude-adjusted value, this would be the less desirable option. Definitely a kludge.

-tk

--

[Drew S.]

To test, I modified line 2928 in ws28xx.py to read "altimeter" instead of "pressure", to see what happens.

You received this message because you are subscribed to a topic in the Google Groups "weewx-user" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/weewx-user/hFp03F9XVH4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to weewx-user+...@googlegroups.com.

[mwall]

On Sunday, May 15, 2016 at 4:12:08 PM UTC-4, Drew S. wrote:

To test, I modified line 2928 in ws28xx.py to read "altimeter" instead of "pressure", to see what happens.

you've got to do it on line 1478 as well.  line 2928 gets the historical records, 1478 gets the LOOP data.

[Les Niles]

Of course if you do #1 then the barometer displayed on the console would not be useful since it would be the uncorrected gauge pressure. 

And there’s also option #3 (gotta have a #3): the driver could undo the offset, then re-adjust correctly.

  -Les

[Drew S.]

That seems to have brought CWOP back into range with surrounding stations: http://www.findu.com/cgi-bin/wxnear.cgi?call=DW8525

But the barometer value is now not being generated for WU or for the local weewx page.

[Drew S.]

Should I undo my changes to ws28xx.py, and instead put something in the [[corrections]] area to "undo" whatever calculation is being done to throw the pressure reading out of whack?

On 05/15/2016 02:08 PM, Thomas Keffer wrote:

You received this message because you are subscribed to a topic in the Google Groups "weewx-user" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/weewx-user/hFp03F9XVH4/unsubscribe.
To unsubscribe from this group and all its topics, send an email to weewx-user+...@googlegroups.com.

[Thomas Keffer]

Great idea. That should work.

-tk

[Drew S.]

Sweet. What should I put in there? I have no clue what calculation is being done, or how.

[Drew S.]

Oh, and regarding choice #1, my console pressure had not been adjusted until a few hours ago, as seen on that graph. Yet it was still reporting too high before the adjustment.

[Thomas Keffer]

Do you recall how big an adjustment?

-tk

[Thomas Keffer]

1. You need a local pressure that you believe in. Pick a local airport. Write down the "altimeter" pressure they are reporting. Write down its altitude.
   
2. At the same time, write down the pressure you're seeing on the console.
   
3. Back calculate the local gauge pressure at the airport. Weewx has a function to do this (weewx.uwxutils.AltimeterToStationPressure()). If you don't know how to call the function, send me the data and I can calculate it.
   
4. Take the difference between #2 and #3. Enter this into [[corrections]].

-tk

[Drew S.]

I think I had to adjust it up about .12-.14 inHg to match the other station in town.

[mwall]

On Sunday, May 15, 2016 at 4:33:10 PM UTC-4, Drew S. wrote:

That seems to have brought CWOP back into range with surrounding stations: http://www.findu.com/cgi-bin/wxnear.cgi?call=DW8525

But the barometer value is now not being generated for WU or for the local weewx page.

from the comments for StdWXCalculate in wxservices.py:

    There is one situation where dependencies matter: pressure.  In the case   
    where the hardware reports barometer, we must calculate pressure and       
    altimeter.  Since altimeter depends on pressure, pressure must be          
    calculated before altimeter.                                               
                                                                               
    We do not handle the situation where hardware reports altimeter and        
    we must calculate barometer and pressure.                                  
 

[Drew S.]

1: KITR 30.00 inHg 4193ft
2: 29.92 inHg

[Drew S.]

Instead of "altimeter", I changed the two lines to "barometer". It seems to be reporting now.

[Thomas Keffer]

The gauge pressure at KITR is 25.74 inHg.

You're seeing 29.92. So, enter a correction of -4.18

-tk

[Thomas Keffer]

Just realized that KITR is at a slightly different altitude than you (4193' vs 4250'). That's an extra -0.05 inHg correction, for a total of -4.23.

-tk

[Drew S.]

I put that in as "pressure = (pressure - 4.18)"?

[Thomas Keffer]

yes. Except use -4.23:

pressure = pressure - 4.23

-tk

[Drew S.]

It seems to be working. I reverted ws28xx.py to the original, and added that line to [[corrections]]. CWOP is staying in range with the other stations, and WU is back down in normal range. Thanks a lot for your help. :)

[Drew S.]

Been tweaking the number lately to try to calibrate it in line with what CWOP expects in their QC. http://weather.gladstonefamily.net/site/D8525 I think I finally got it dialled in with -4.08.

[Thomas Keffer]

The CWOP expectations are modeled. I would not treat them as gospel. You're better off finding a calibrated barometer and tuning to that.

-tk