Help me understand altimeter difference on CWOP

[Nate Bargmann]

My apologies for hitting what is probably an FAQ and a sore spot.

I have had my Vantage Pro2 online for a bit over a week and using WeeWX
which has been working very well. As I had been uploading to Weather
Underground for about a decade it was the first service I made certain
was receiving data. Almost immediately the barometer reading on the
console was about 0.05 inHg below the two nearest ASOS airport readings
so I naturally added an offset using the console and WU is satisfied
with my data.

A week ago yesterday I began submitting data to APRS-IS and immediately
noticed that my data was running 2, sometimes 3 mBar below the local
airport readings. I am now set up in MADIS and today received my first
data quality report. As I expected from watching this over the past
week the barometric data is running low and the Weather Quality report
appears to confirm this by showing the barometer for 21 May being 2.23
mBar low.

I am puzzled as the local airports METAR show the pressure as being the
altimeter since it is prefixed with an 'A':

METAR KBIE 222335Z AUTO 14008KT 10SM OVC060 17/06 A3017 RMK AO2 T01650059
METAR KMYZ 222315Z AUTO 14009KT 10SM OVC060 16/09 A3017 RMK AO2

I've yet to see SLP in these reports.

At the same time, my station report this to CWOP:

N0NB>APRS,TCPIP*,qAC,SIXTH:@222336z3954.68N/09646.27W_112/008g010t062r000p000P000b10190h66.weewx-3.1.0-Vantage

Per a chart I found online, 1019.0 mBar converts to 30.09 inHg, a
significant difference as I am 8 miles from KMYZ and about 35 feet
higher in ground elevation.

At the moment, the nearby CWOP activity shows my pressure being lower
than the surrounding stations (never mind the two readings in the 978
mBar range):

http://www.findu.com/cgi-bin/wxnear.cgi?call=N0NB

Finally, the MADIS quality check shows my station being more than twice
the standard deviation at 2.2 milliBars:

http://weather.gladstonefamily.net/site/AV057?days=3#Data

I have both WeeWX and the Vantage console set to the same elevation of
1327 feet, which is the console elevation. Ground elevation is 1319
feet (I guess I'm entered into MADIS at 1318 feet).

I find it interesting that the closest CWOP station is reporting
pressure values in line with the nearest airport readings. Yes, I
realize there is a difference in pressure readings and, as I understand
it, this issue is complicated by the Davis console not providing the raw
station pressure. I've been reading quite a lot the past several days
from various sources on the Web. :-)

The reason for my post is, where do I go from here? I do not wish to
blow up WeeWX if it is working fine on CWOP for everyone else. I would
like to provide readings to CWOP that are in line with what MADIS
considers good quality data and that are in line with the local airport
readings.

Thanks for reading.

- Nate

--

"The optimist proclaims that we live in the best of all
possible worlds. The pessimist fears this is true."

Ham radio, Linux, bikes, and more: http://www.n0nb.us

[Thomas Keffer]

Weewx uses altimeter pressure when reporting to CWOP.

There are many, many discussions about this a few years ago. If you try searching the group, you can find several threads about this.

There's also a bit of discussion in the Customizing Guide, under writing a driver.

-tk

--
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.

[Nate Bargmann]

Thanks for the reply, Tom.

* On 2015 22 May 21:20 -0500, Thomas Keffer wrote:
> Weewx uses altimeter pressure when reporting to CWOP.

I understand this. What I don't understand is that if the airport
stations referenced by NWS are also reporting altimeter, why doesn't my
station match their value? Is this a failure of my console? Is there a
way to see the internal pressure value (this is a brand new Vantage Pro2 console).

> There are many, many discussions about this a few years ago. If you try
> searching the group, you can find several threads about this.

Yes, I read most of them as I found them over the past week.

> There's also a bit of discussion in the Customizing Guide, under writing a

> driver <http://weewx.com/docs/customizing.htm#genLoopPackets>.

I had read that as well and it is a very good description. I have no
argument against it.

What I am not understanding that since WeeWx is reporting the correct
value why MADIS says it is in error.

From reading the archives I know a lot of thought and work went into
this and I am not asking to reopen those discussions. It's that I
suspect that if I were to bring this up on the WXQC mailing list that I
would be told that my station is broken. That is not a useful answer
for any of us.

At some point in time I will likely be asked about this apparent data
error and I want to be prepared to defend my data and by extension
WeeWX.

[Liz]

On Fri, 22 May 2015 22:48:13 -0500
Nate Bargmann <n0...@n0nb.us> wrote:

> What I am not understanding that since WeeWx is reporting the correct
> value why MADIS says it is in error.
>
> From reading the archives I know a lot of thought and work went into
> this and I am not asking to reopen those discussions. It's that I
> suspect that if I were to bring this up on the WXQC mailing list that
> I would be told that my station is broken. That is not a useful
> answer for any of us.

Objectively, there are two data sources, and they both have inherent
errors, and so the pressures are X +/- x and Y +/- y.
Both stations could be reporting pressure which is correct within
published tolerances. An assumption has been made that the airport
station is correct, and a second assumption that yours is incorrect.
How often is the airport station calibrated? What is its drift over
time?
You have advised that yours is new, and so I assume that its
calibration has been recently checked.

In a different country to you, and I have seen my airport station
several degrees out on the temperature, compared to 3 other stations
within 10km. I don't compare the pressure readings across the stations,
but I expect to see variation related to local wind, updrafts and so on.

Liz

[Nate Bargmann]

* On 2015 23 May 01:04 -0500, Liz wrote:

> Objectively, there are two data sources, and they both have inherent
> errors, and so the pressures are X +/- x and Y +/- y.

The two sources being the local airport and my station, correct? Or do
you mean both local airports?

> Both stations could be reporting pressure which is correct within
> published tolerances. An assumption has been made that the airport
> station is correct, and a second assumption that yours is incorrect.
> How often is the airport station calibrated? What is its drift over
> time?

Of course the assumption will be that the private station is incorrect
since it is consumer grade equipment and only factory calibrated. As to
when the airport equipment is calibrated, I have no idea, however, going
out a bit wider and adding three more airports from the region that are
with 60 miles, casually watching the trends over the past decade has not
revealed any glaring errors over that period of time. To be fair the
local airport, KMYZ, has only had its data available within the past
year that I am aware of.

> You have advised that yours is new, and so I assume that its
> calibration has been recently checked.

It would have whatever calibration is done at the factory. Of course, I
can change the console reading to whatever I like, within some range I
presume, to adjust to local reporting. What I have failed to do is keep
a log of my adjustments so I don't know exactly how much I have changed
(what Davis calls calibration) the reading. :-(

Perhaps I need to reset the console to factory defaults and record
things more carefully.

> In a different country to you, and I have seen my airport station
> several degrees out on the temperature, compared to 3 other stations
> within 10km. I don't compare the pressure readings across the stations,
> but I expect to see variation related to local wind, updrafts and so
> on.

And so far, despite the apparent error, my barometer does seem to track
variations reliably which may be of more value than absolute readings.

On a related note, comparing the wxservices.py file from the GitHub
repository to the version in 3.1.0 that is in use on my Pi, I see a
number of changes have been made as well as changes to the config file.
At this time it may be better for me to await the release of 3.2 and see
what changes my data receive at that time.

- Nate -- anxiously awaiting WeeWX 3.2.0 :-)

[Chip Cuccio]

> On May 23, 2015, at 9:04 AM, Nate Bargmann <n0...@n0nb.us> wrote:

Nate,

If you’re not already, you may want to consider subscribing to the CWOP (and MADIS)-centric WXQC mailing list:
<http://server.gladstonefamily.net/mailman/listinfo/wxqc>

Sharp people on that list with excellent discussions, and the topic such as yours is a frequent one.

—
Chip

[Nate Bargmann]

* On 2015 24 May 13:12 -0500, Chip Cuccio wrote:

> If you’re not already, you may want to consider subscribing to the
> CWOP (and MADIS)-centric WXQC mailing list:
> <http://server.gladstonefamily.net/mailman/listinfo/wxqc>

Yes, I did subscribe a few days ago. Thanks, Chip.

> Sharp people on that list with excellent discussions, and the topic
> such as yours is a frequent one.

That's partly why I am reluctant to ask since this is an FAQ as it's
probable that the answers will suspect weeWX to be at fault. Right now
my console is showing 29.80 inHg and weeWX is reporting that value to
Weather Underground while sending 1008.1 mBar to CWOP which Google
converts to 29.77 inHg. 17 minutes ago the closest four airports all
reported 29.83 inHg. Barometers are falling so I probably shouldn't go
to tweaking my calibration setting.

- Nate

[Nate Bargmann]

Forgive me as this still is bugging me. :-D

I've been pondering this and am observing the following.

1) WeeWX expects a value of "barometer" from the Vantage console that is
reflective of Sea Level Pressure.

2) From this value WeeWX sets about to calculate Station Pressure and
Altimeter.

Given the two conditions above, unless the value obtained from the
Vantage logger archive record is SLP, nothing else will be correct.

Given that the local airports only provide altimeter values in their
METAR reports and do not report SLP, setting my "barometer" on the
console to match the airport altimeter reports results in errors down
the line.

To correct this, it seems to me that I need to adjust the "barometer"
setting on my console higher until the altimeter value that WeeWX
reports to CWOP matches the values reported by the closest airports.
This also means that I should make a change so that WeeWX also reports
the altimeter value to Weather Underground instead of "barometer". Or,
won't WU care?

Once I cease putting garbage into WeeWX, the data quality will improve.
:-)

[Thomas Keffer]

Your analysis is pretty much correct: you cannot compare the altitude and temperature corrected SLP pressure displayed on the console of the Vantage with the "altimeter" pressure reported by the local airport. You'll need to "correct" one or the other. Your iterative process is as good as any for doing this.

Wunderground does not say whether they expect altitude and temperature corrected, or just altitude corrected, pressure. Weewx uses the former when reporting to WU.

-tk

[Nate Bargmann]

* On 2015 31 May 09:18 -0500, Thomas Keffer wrote:
> Your analysis is pretty much correct: you cannot compare the altitude and
> temperature corrected SLP pressure displayed on the console of the Vantage
> with the "altimeter" pressure reported by the local airport. You'll need to
> "correct" one or the other. Your iterative process is as good as any for
> doing this.
>
> Wunderground does not say whether they expect altitude and temperature
> corrected, or just altitude corrected, pressure

> <http://wiki.wunderground.com/index.php/PWS_-_Upload_Protocol>. Weewx uses

> the former when reporting to WU.

Ahh, the joy of ambiguity, eh? :-D

Another thought I had and wanted to wait until I received your feedback
is this. I have read in various places on the Web where Davis states
that setting the console elevation value to 0 will result in the raw
station pressure reading in inHG being reported. How does WeeWX handle
this when the actual elevation is given in weewx.conf?

For example, set station console elevation to 0, set 'altitude' in
weewx.conf to 1327, foot.

As my system is live and MADIS analysis is now looking good, I've been
reluctant to try this idea. :-)

[Thomas Keffer]

If the hardware supports it, weewx always gets altitude off the console and ignores the value in weewx.conf. That would be the case with a Vantage.

-tk