CWOP and barometer reading

[Super-T]

I currently have EWpressureoffset=10.5 and this gives me a reading very close to the local Airport readings but not near the reading worked by formula.
Maybe my height is wrong as I used the reading from Google earth. How do other people work out height above sea level?
If an aircraft was using the reading at CWOP then it would probably be buried 10 metres into my backyard.

Local APRS Stations http://www.findu.com/cgi-bin/wxnear.cgi?call=DW1926

[Super-T]

Rejester. Where do you do the MADIS thumbs test?

[steve]

Maybe my height is wrong as I used the reading from Google earth. How do other people work out height above sea level?

If the result of the calculation is too low compared to your local airfield, then you probably have your altitude set too low. Or my calculation is wrong

We have Ordnance Survey maps here, which are very good at allowing you to find your altitude to a good level of accuracy. Also available for free (a small section at a time) on the internet. Nothing similar in NZ?

Steve

[rejester]

My baro is right on now (according to MADIS), but I still plan to load the beta update tonight and see if it changes. It takes a day or so to get a reading back from them.

Terry - Since I'm reporting to CWOP, an option is to get quality control reports from MADIS. I signed up. The link to my reports are here: http://weather.gladstonefamily.net/site/C0945
Signup information is here: http://www.wxqa.com/aprswxnetqc.html

And actually, my baro reading is going south again....I just checked it. As of yesterday morning it was doing better. This will be a good time for me to see if the update does anything for me.

Ryan

[rejester]

Well, prior to 792 my barometer was reading higher than the computed value (what MADIS thinks it should be). After the 792 it's reading about the same error amt but now lower than the computed value. Typical amount I'm off, when it's off, is 2.5 to 3.5 mb. Like I mentioned before, sometimes it's right on so changing a calibration offset won't do any good.

Ryan

[steve]

Because of the complicated way the value is derived in Cumulus for VP stations, I think I'm going to have to put some extra diagnostics in to see where the calculations are going wrong (I know, I should have put them in any way, on the assumption there would be problems!). It first has to derive the station pressure and then calculate the altimeter pressure. Either or both could be going wrong. I'm going to be uploading a new build later today, I'll add the diagnostics to that.

Steve

[steve]

Ryan, build 794 has some extra diagnostics (just logging the parameters that go into the calculations) for the VP altimeter pressure calculation. Turn on the debug log, run for a few minutes (at least 5), and then let me have the debug.log file, please.

Steve

[rejester]

Steve,

Build 794 running.
The debug log is attached, 8 min or so. Passwords removed. (side note, any way to not pass those as plain text to the log?)
Thanks for checking it out.

Ryan

[steve]

There was an error in my tracing code, so I'm tracing the elevation twice instead of the humidity correction. I've run a corrected version here, and the figures do seem to be correct. I'm practically at sea level (elevation 16 ft) so you could argue it's not a true test. But I'm using Steve Hatchett's code to do the calculations, and I've checked that I haven't screwed up the numbers in my own code, by writing a little test app that only uses Steve's code, with none of the unit conversions that Cumulus does, and the Cumulus numbers agree with the test program numbers.

So I don't know where we go from here; either the numbers are correct, or Steve's code is wrong (which I think is highly unlikely).

I did notice that Cumulus may be introducing some very small error in its unit conversions; your altimeter pressure should have been reported as 1014.6 at 10:33, and Cumulus sent 1014.5, but this is a 0.01% error, so I don't think this is relevant.

I suppose that you could run with the corrected diagnostics so we at least have all the input data.

Edit: Actually, I suppose another possibility is that sometimes Cumulus is getting the wrong calibration data from the station...

Steve

[steve]

I've fixed the tracing in the latest build. I guess what we really need is the trace from a time when your pressure is being flagged as wrong.

Steve

[rejester]

I'll try to get some of my data from CWOP and see when things go awry. I think I've put you through enough exercise with this, it could be an issue with my equipment. The fact no one else is chiming in makes me wonder, or perhaps not many report to CWOP, or maybe they are not doing quality checks. I do however feel more confident in the way pressure is being sent to CWOP, thanks for changing that.

Ryan

[M1DUL]

I'll try to get some of my data from CWOP and see when things go awry. I think I've put you through enough exercise with this, it could be an issue with my equipment. The fact no one else is chiming in makes me wonder, or perhaps not many report to CWOP, or maybe they are not doing quality checks. I do however feel more confident in the way pressure is being sent to CWOP, thanks for changing that.

Ryan

Is this the website you do the quality checks on?

http://weather.gladstonefamily.net/site/D2423 if it is I am also getting the same problem.

[rejester]

Michael, yes, that's the same site I'm using.

Ryan

[steve]

I am also getting the same problem.

As you have a Fine Offset station, the calculation is considerably simpler, as the station supplies the station pressure and Cumulus doesn't have to try to work it out. It just has to calculate the altimeter pressure from the station pressure, and it does this with the algorithm I quoted earlier, from Steve Hatchett:

Code: Select all

k1 := 0.190284; // discrepency with calculated k1 probably because Smithsonian used less precise gas constant and gravity values
k2 := 8.4184960528E-5; // (standardLapseRate / standardTempK) * (Power(standardSLP, k1)
Result := Power(Power(pressureHPa - 0.3, k1) + (k2 * elevationM), 1/k1);

Where pressureHPa is the station (absolute) pressure read from the station, and elevationM is your elevation in metres as configured in Cumulus (converted from feet if necessary).

So, if your CWOP data is being flagged as wrong (but I can't see it being flagged as wrong on that page?), then either your station is supplying an incorrect value, or your elevation in Cumulus is wrong.

Steve

[M1DUL]

Thank you for the Info Steve, I did change the hight in Cumulus last night and it looks as though it is all right now.

I am also getting the same problem.

As you have a Fine Offset station, the calculation is considerably simpler, as the station supplies the station pressure and Cumulus doesn't have to try to work it out. It just has to calculate the altimeter pressure from the station pressure, and it does this with the algorithm I quoted earlier, from Steve Hatchett:

Code: Select all

k1 := 0.190284; // discrepency with calculated k1 probably because Smithsonian used less precise gas constant and gravity values
k2 := 8.4184960528E-5; // (standardLapseRate / standardTempK) * (Power(standardSLP, k1)
Result := Power(Power(pressureHPa - 0.3, k1) + (k2 * elevationM), 1/k1);

Where pressureHPa is the station (absolute) pressure read from the station, and elevationM is your elevation in metres as configured in Cumulus (converted from feet if necessary).

So, if your CWOP data is being flagged as wrong (but I can't see it being flagged as wrong on that page?), then either your station is supplying an incorrect value, or your elevation in Cumulus is wrong.

Steve