Topic: add pws address / elevation problems

[farcus]

adding a new pws in WU - if using the address finder and selecting my address it now pinpoints on the map a location about 200 metres away. This seems to be a new issue since the updates as it was always pinpoint accurate on the map previously.

Also, whether adding by address or manually on the map elevation is shown correctly.
However, after adding and the pws is showing as online, in the dashboard elevation is shown as 0 mt / 0 ft

 [ You are not allowed to view attachments ]

 [ You are not allowed to view attachments ]

[WSWeather]

It appears they've degraded the geolocation accuracy considerably, down to about +/-36 seconds.  Depending on where you are on the planet that can be as much as half a mile (800 meters) or more.  It is affecting a lot of things so you're probably better off trying to find your location on Google Maps and use the "what's here" function to get the coordinates and enter them directly.

[farcus]

It appears they've degraded the geolocation accuracy considerably, down to about +/-36 seconds.  Depending on where you are on the planet that can be as much as half a mile (800 meters) or more.  It is affecting a lot of things so you're probably better off trying to find your location on Google Maps and use the "what's here" function to get the coordinates and enter them directly.

in the screenshot above the coordinates are correct and match that of google maps. This is from position the pointer manually rather than typing in the address.
The only thing incorrect there is the elevation.

[galfert]

Yeah but what WSWeahter is saying is that with only 3 decimal Degrees there is not enough precision. Except he was off by a decimal place. Precision is off by +/- 3.6 seconds not 36. But that still translates to 111.32 meters or 365.2 feet. I would like to see 6 decimal places used as most other weather software does. At least 5 decimal places is warranted...but I really want to see 6.

See chart
https://en.m.wikipedia.org/wiki/Decimal_degrees

[farcus]

the elevation issues appear to be fixed now - although I had to re-add my station to get it to show the correct (or any) elevation.
I tried using Google coordinates when adding (6 decimal places) but WU just kept asking for valid coordinates until I reduced to 3 decimal places. 

[galfert]

Strange. I just created a new dummy station and I started by clicking on the map...then I edited the location and added more digits to the 3 decimal numbers shown. I took it to 6 decimals and it let me create the station. Interestingly though when I went to My Devices and then clicked to edit that newly created station it shows it to me with 8 decimals precision. Perhaps all places on the map are not treated equally. Dunno what else to say.

[Mattk]

Ok so what I am missing here? 

In original post position Glenwood is 44.39°S, 171.21°E which would be in NZ, then in the Personal Weather Station info box Glenwood position -44.387°N, 171.212°W which is garbage, but then what would WU know? 

[Mattk]

Yeah but what WSWeahter is saying is that with only 3 decimal Degrees there is not enough precision. Except he was off by a decimal place. Precision is off by +/- 3.6 seconds not 36. But that still translates to 111.32 meters or 365.2 feet. I would like to see 6 decimal places used as most other weather software does. At least 5 decimal places is warranted...but I really want to see 6.

111.32m is rather irrelevant as that would be at the equator and as you are high on so called precision then for the record for the location in question +/- 3.6" would be 111.1199m (yes that's precise) if referring to Latitude, 79.6789m if referring to Longitude but as positions are at least 2D then could also be 136.7342m

6 decimal places of a degree I think you are having yourself on as I would doubt you have the capability to determine a location to less than 0.15m (6"), 5 decimal places I would say you are still kidding yourself which is around 1.4m, honestly what's the point wanting 1.4m, you also probably don't have the capability to achieve that accuracy either??     

 

[galfert]

Yeah but what WSWeahter is saying is that with only 3 decimal Degrees there is not enough precision. Except he was off by a decimal place. Precision is off by +/- 3.6 seconds not 36. But that still translates to 111.32 meters or 365.2 feet. I would like to see 6 decimal places used as most other weather software does. At least 5 decimal places is warranted...but I really want to see 6.

111.32m is rather irrelevant as that would be at the equator and as you are high on so called precision then for the record for the location in question +/- 3.6" would be 111.1199m (yes that's precise) if referring to Latitude, 79.6789m if referring to Longitude but as positions are at least 2D then could also be 136.7342m

6 decimal places of a degree I think you are having yourself on as I would doubt you have the capability to determine a location to less than 0.15m (6"), 5 decimal places I would say you are still kidding yourself which is around 1.4m, honestly what's the point wanting 1.4m, you also probably don't have the capability to achieve that accuracy either??     

 

https://gizmodo.com/a-new-technique-makes-gps-accurate-to-an-inch-1758457807

[Mattk]

Yeah but what WSWeahter is saying is that with only 3 decimal Degrees there is not enough precision. Except he was off by a decimal place. Precision is off by +/- 3.6 seconds not 36. But that still translates to 111.32 meters or 365.2 feet. I would like to see 6 decimal places used as most other weather software does. At least 5 decimal places is warranted...but I really want to see 6.

111.32m is rather irrelevant as that would be at the equator and as you are high on so called precision then for the record for the location in question +/- 3.6" would be 111.1199m (yes that's precise) if referring to Latitude, 79.6789m if referring to Longitude but as positions are at least 2D then could also be 136.7342m

6 decimal places of a degree I think you are having yourself on as I would doubt you have the capability to determine a location to less than 0.15m (6"), 5 decimal places I would say you are still kidding yourself which is around 1.4m, honestly what's the point wanting 1.4m, you also probably don't have the capability to achieve that accuracy either??     

 

https://gizmodo.com/a-new-technique-makes-gps-accurate-to-an-inch-1758457807

Well isn't that a wishy washy link that basically says jack. We've had the ability for cm type accuracy for the past 25 years but that isn't something you can probably do even now as the system simply can not achieve that type of accuracy without augmentation and if you really understood GPS then you wouldn't be poking through links like that.

Like you could get cm accuracy now but like what's the point of the exercise with a weather station (which is the real issue here), will knowing the location of a weather station to 1cm or 15cm or 1.5m or 10m achieve what? 10 metres is quite achievable but still rather irrelevant to a weather station.   

[galfert]

Yeah but what WSWeahter is saying is that with only 3 decimal Degrees there is not enough precision. Except he was off by a decimal place. Precision is off by +/- 3.6 seconds not 36. But that still translates to 111.32 meters or 365.2 feet. I would like to see 6 decimal places used as most other weather software does. At least 5 decimal places is warranted...but I really want to see 6.

111.32m is rather irrelevant as that would be at the equator and as you are high on so called precision then for the record for the location in question +/- 3.6" would be 111.1199m (yes that's precise) if referring to Latitude, 79.6789m if referring to Longitude but as positions are at least 2D then could also be 136.7342m

6 decimal places of a degree I think you are having yourself on as I would doubt you have the capability to determine a location to less than 0.15m (6"), 5 decimal places I would say you are still kidding yourself which is around 1.4m, honestly what's the point wanting 1.4m, you also probably don't have the capability to achieve that accuracy either??     

 

https://gizmodo.com/a-new-technique-makes-gps-accurate-to-an-inch-1758457807

Well isn't that a wishy washy link that basically says jack. We've had the ability for cm type accuracy for the past 25 years but that isn't something you can probably do even now as the system simply can not achieve that type of accuracy without augmentation and if you really understood GPS then you wouldn't be poking through links like that.

Like you could get cm accuracy now but like what's the point of the exercise with a weather station (which is the real issue here), will knowing the location of a weather station to 1cm or 15cm or 1.5m or 10m achieve what? 10 metres is quite achievable but still rather irrelevant to a weather station.

You can also get some pretty good accuracy using online maps with satellite imaging. Just point and click and there you have a very accurate location that you can visually see. The point of this exercise to to make good on making things as precise as possible. You take the same care with your barometer and other sensors....so why not apply the same to your location. It doesn't do that much in the grand scheme of things but you know you've got a well documented and sited station. Nobody is getting illusions of feeling like they achieved some miraculous measurement precision. It is a matter of consistency also across software and websites. If they all stuck to 6 decimal places it would go over well with a lot of people. If I say my station is at 42.123456, -110.123456 then that is where it is and I don't think anyone is going to lose sleep over its precision....but having a consistent measurement goes a long way. I don't want to sometimes have to round to 5 digits and other times have different software ask me to round to 4 digits. There is no harm in just using 6 digits even if for all practical purposes it is difficult to get that accuracy.

[Mattk]

...If I say my station is at 42.123456, -110.123456 then that is where it is ...

  Unless you have the capability of actually fixing that position then you really don't know exactly where it is and the justification for what is unachievable doesn't stack up practically no matter what you say it is.

Not that it is relevant but since you are all over so called precision and bring up GPS accuracy claims written by basically amateurs it's interesting that GPS only uses 14 decimal places for Pi and to maintain the system accuracy must only use 14 decimal places as in your logic 16 decimal places would be more accurate (so called) but in fact destroys the whole integrity of the system. Some useless info for today

[konz]

Yeah but what WSWeahter is saying is that with only 3 decimal Degrees there is not enough precision. Except he was off by a decimal place. Precision is off by +/- 3.6 seconds not 36. But that still translates to 111.32 meters or 365.2 feet. I would like to see 6 decimal places used as most other weather software does. At least 5 decimal places is warranted...but I really want to see 6.

111.32m is rather irrelevant as that would be at the equator and as you are high on so called precision then for the record for the location in question +/- 3.6" would be 111.1199m (yes that's precise) if referring to Latitude, 79.6789m if referring to Longitude but as positions are at least 2D then could also be 136.7342m

6 decimal places of a degree I think you are having yourself on as I would doubt you have the capability to determine a location to less than 0.15m (6"), 5 decimal places I would say you are still kidding yourself which is around 1.4m, honestly what's the point wanting 1.4m, you also probably don't have the capability to achieve that accuracy either??     

 

https://gizmodo.com/a-new-technique-makes-gps-accurate-to-an-inch-1758457807

Well isn't that a wishy washy link that basically says jack. We've had the ability for cm type accuracy for the past 25 years but that isn't something you can probably do even now as the system simply can not achieve that type of accuracy without augmentation and if you really understood GPS then you wouldn't be poking through links like that.

Like you could get cm accuracy now but like what's the point of the exercise with a weather station (which is the real issue here), will knowing the location of a weather station to 1cm or 15cm or 1.5m or 10m achieve what? 10 metres is quite achievable but still rather irrelevant to a weather station.

You can also get some pretty good accuracy using online maps with satellite imaging. Just point and click and there you have a very accurate location that you can visually see. The point of this exercise to to make good on making things as precise as possible....[snip]

Um, no.  Satellite and other imagery is only as good as the network used to tie it to the ground.  Imagery doesn't just come out of the camera accurate.  Google imagery can be meters off in some locations.  It is [or, used to be] ultra expensive to get <1ft horizontal accuracy (x,y) on imagery, let alone vertical accuracy (z).  Think of it as a photographic approximation of the earth's surface.  Clicking on a map where the manhole is in front of your house probably won't yield you an accurate coordinate or elevation, ever.  A licensed surveyor is probably required for this.

However, I do agree with the relative accuracy parts of your post.   

[galfert]

You all are missing my point which I already said. It's not as much about precision as it is about a uniform experience across applications. I recently set up a station with Ecowitt.net and clicking on their map yielded a location on the map with 13 decimal places. That is ridiculous and unwarranted...ever! All I'm saying is that 6 decimal places sounds like a good number that I can use everywhere and feel comfortable about not being tedious to remember and not having to round down to 3 or 4 or 5 decimal places depending on what the software of the day feels like using.

Click on a location with Google Maps and what do you get? ...6 decimal places. I like sticking to what Google says.

[Mattk]

You all are missing my point which I already said. It's not as much about precision as it is about a uniform experience across applications. I recently set up a station with Ecowitt.net and clicking on their map yielded a location on the map with 13 decimal places. That is ridiculous and unwarranted...ever! All I'm saying is that 6 decimal places sounds like a good number that I can use everywhere and feel comfortable about not being tedious to remember and not having to round down to 3 or 4 or 5 decimal places depending on what the software of the day feels like using.

Yes somebody is missing the point here, entirely. There's no way known you are going to get anything better than map scale accuracy by clicking on a map. A modern era version of calculator syndrome. Systems with 13 decimals or basically anything above about 4 or 5 are clearly showing exactly what they know about positioning in the real world for the purpose, and that's not a lot when it's appears nothing more about feel good.

But still there have been absolutely no valid explanation for 6 decimal places in a weather station location, none what so ever. How can anybody actually round down to 4 or 5 places when the starting value is basically rubbish and not achievable (at this level) in the first place?   

[Mattk]

You all are missing my point which I already said. It's not as much about precision as it is about a uniform experience across applications. I recently set up a station with Ecowitt.net and clicking on their map yielded a location on the map with 13 decimal places. That is ridiculous and unwarranted...ever! All I'm saying is that 6 decimal places sounds like a good number that I can use everywhere and feel comfortable about not being tedious to remember and not having to round down to 3 or 4 or 5 decimal places depending on what the software of the day feels like using.

Click on a location with Google Maps and what do you get? ...6 decimal places. I like sticking to what Google says.

Google syndrome, they simply don't have that accuracy, talk about a con job, 6 decimal places out of Google is absolutely rubbish

[galfert]

Well you had your point and I had mine.

Cheers 

[Mattk]

Then you will have to live in your own pretend dream world .

[Mattk]

Ok so what I am missing here? 

In original post position Glenwood is 44.39°S, 171.21°E which would be in NZ, then in the Personal Weather Station info box Glenwood position -44.387°N, 171.212°W which is garbage, but then what would WU know?

How about this one? Anybody have an explanation for this one?

[farcus]

actually, the info in Personal Station info is probably the accurate one.
Here is the info with my station re-added. The highlighted in red seems to be the garbage info.
 [ You are not allowed to view attachments ]

just to confirm . . . the correct elevation is 52 meters or 171 ft.

[farcus]

and this is what the old dashboard is currently showing . . .

 [ You are not allowed to view attachments ]

 [ You are not allowed to view attachments ]

[Mattk]

actually, the info in Personal Station info is probably the accurate one.
Here is the info with my station re-added. The highlighted in red seems to be the garbage info.

The lat/Long in the red box appears fine, the Longitude of 171.212°W is well to the east of NZ on the other side of the western hemisphere

[Tizme]

I have a similar problem with the elevation on the main page of NZ station IWHAKATA11 showing "1161 mt" instead of correctly in metric as 353m as shown in the Station detail panel.

I have the Lat / Long correctly listed on the main screen (38.02 S  176.64 E)   and incorrectly in the station detail screen - again way out in the Pacific ocean at -38.02S  176.637 W.
A couple of errors that seem pretty consistant for a number of stations I checked.
No doubt it will all come out in the wash if the forecast is correct.

[Mattk]

I have a similar problem with the elevation on the main page of NZ station IWHAKATA11 showing "1161 mt" instead of correctly in metric as 353m as shown in the Station detail panel.

I have the Lat / Long correctly listed on the main screen (38.02 S  176.64 E)   and incorrectly in the station detail screen - again way out in the Pacific ocean at -38.02S  176.637 W.
A couple of errors that seem pretty consistant for a number of stations I checked.
No doubt it will all come out in the wash if the forecast is correct.

Yes appears to be consistent with WU having a problem with their positions and heights? Does this surprise anyone?

353 metres is 1158 feet, which is ball park with 1161 but if "mt" is supposed to represent MeTres then obviously there is some confusion between Metres & Feet?

-38.02S  176.637 W are again nonsense values which doesn't represent the actual station position in any language.