Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

GPS accuracy question

  • Please log in to reply
19 replies to this topic

#1 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 05 April 2024 - 10:50 PM

I have a small GPS device (Dual XGPS), and today I used it to measure the coordinates of my observing site.  Then I went to Xavier's interactive eclipse map and click at the location on the map.  The interactive map would show the coordinates as well.

 

The issue is, the coordinates do not match.  The longitude shown by the GPS device (on its iPhone APP) is a whole 1' west to the coordinate shown on Xavier's interactive map.  1' is a whole lot.  Something like several hundreds times larger than the possible error of my click on the map. It can lead to 1 to 2 seconds of time difference for C2 to C3 in SEM.

 

So, I wonder which one I should trust.  The hardware GPS?  Or the map?

 

It's also a bit strange that the offset is exactly 1'.  Not 1'15", not 56", not 1'06".  It's 1'00".  If it's someone's handwriting, I would say it's a typo of a single digit among those many digits.  But it's not handwriting.



#2 chvvkumar

chvvkumar

    Surveyor 1

  • *****
  • Posts: 1,822
  • Joined: 03 Aug 2020
  • Loc: Cottleville, MO

Posted 05 April 2024 - 11:26 PM

If the GPS had a good satellite lock for some duration and you know the GPS is good, I would trust the GPS over maps. 



#3 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 05 April 2024 - 11:30 PM

Thanks.  I find no reason not to trust the GPS.  But I don't know if there are good reasons not to trust the map.  Can the map be this wrong?



#4 Anhydrite

Anhydrite

    Ranger 4

  • -----
  • Posts: 356
  • Joined: 26 Aug 2017

Posted 06 April 2024 - 12:22 AM

Were the different display formats

 

like Degrees Minutes Seconds, or degrees decimal minutes?



#5 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 06 April 2024 - 12:39 AM

Same format.  One is 106° 26' 33.3" W. The other is 106° 25' 33.3" W.

 

After I check my GPS screenshot, I found that it says the altitude is 22m.  The beach/ocean is right in front of my eyes.  I can't believe I am 22m above sea level.  Does this mean I should trust the GPS reading less?  I know GPS typically have larger altitude errors.  So a wrong altitude does not necessarily imply wrong horizontal coordinates.



#6 Anhydrite

Anhydrite

    Ranger 4

  • -----
  • Posts: 356
  • Joined: 26 Aug 2017

Posted 06 April 2024 - 03:12 AM

I would say that the coords you pull from the map are more accurate.  Maps have been accurate for a long  while.  The GPS may not have a solid solution yet.

 

Did you let it sit for a while after it grabbed all the satelites?



#7 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 06 April 2024 - 09:48 AM

How long should I let it sit?  First time I turned it on, it got more than a dozen of satellites, but cannot offer a solution for 2 minutes.  Then I turn it off and on again.   It provided a solution in a minute and then the value doesn't seem to change much in another minute.  Is this long enough?  Or I should wait for longer?



#8 bookemdano

bookemdano

    Explorer 1

  • *****
  • Posts: 58
  • Joined: 05 Jan 2009

Posted 06 April 2024 - 10:12 AM

Not sure if you are still/already on site for your viewing location but seems to me you can/should be testing this wherever you happen to be right now. Take the GPS unit outside and power it on. Leave it in one place and check the position after 5 minutes, 10 minutes, 20 minutes. See if it changes in that time. If not then it's likely a good fix. Then take those coordinates to Xavier's map and see if they line up.

 

Xavier's maps are incredible. He is one of the world's foremost experts on eclipses so would not expect his map to be wrong to any significant degree.

 

But of course GPS is very accurate as well with a good 3D fix, so I would fully expect the two to correlate. You can also check the GPS coordinates from your smart phone and see if they match.



#9 Chris Johnson

Chris Johnson

    Apollo

  • *****
  • Posts: 1,114
  • Joined: 08 Sep 2004
  • Loc: Stockholm, NJ

Posted 06 April 2024 - 10:45 AM

Use Google earth to find the coordinates of the viewing site. Tap settings and active “Show grid lines” you will then have the lt/long lines. You can also use thr longitude lines to find true nort.


Edited by Chris Johnson, 06 April 2024 - 10:45 AM.


#10 geovermont

geovermont

    Apollo

  • *****
  • Posts: 1,051
  • Joined: 06 May 2016
  • Loc: Vermont

Posted 06 April 2024 - 11:50 AM

This sounds like a tough one to sort out. I've worked as a geologist and cartographer and surveyor for 40 years and have used GPS receivers of all levels of accuracy since the beginning. However, I have no idea how good that app is or how they do their processing. If you used a standard recreation grade receiver with 6 or more satellites in good arrangement and allowed the receiver to run long enough to update the ephemeris, then an error of one minute in lat or long would be very, very unlikely. Maybe a few tenth of seconds of lat/long in bad multipathing situations, but not a whole minute. And although the map is probably great, the person above who expressed unlimited confidence in any map is way off the mark. Flawed maps are put up online all the time. A more likely problem is that the two position sources might be in different datums--there are many, many "flavors" of coordinates out there and this is a very common reason for two sources to disagree. And I would suggest that turning a GPS receiver on and off repeatedly to try to get a good fix is very counterproductive. Leave it running.

 

Good luck!


  • whwang likes this

#11 BlueMoon

BlueMoon

    Aurora

  • *****
  • Posts: 4,691
  • Joined: 14 Jun 2007
  • Loc: South Central Idaho

Posted 06 April 2024 - 11:57 AM

I've found this site to be fairly accurate: https://latitude.to/ and it lists coordinates in DD, DMS, UTC and Geohash formats.

 

I use an Android polar alignment app and my AZMP mount reads GPS signal. The AZMP and the website agree with one another. The app is about 1 arc min off. Cheers.


Edited by BlueMoon, 06 April 2024 - 01:30 PM.


#12 bookemdano

bookemdano

    Explorer 1

  • *****
  • Posts: 58
  • Joined: 05 Jan 2009

Posted 06 April 2024 - 12:31 PM

This sounds like a tough one to sort out. I've worked as a geologist and cartographer and surveyor for 40 years and have used GPS receivers of all levels of accuracy since the beginning. However, I have no idea how good that app is or how they do their processing. If you used a standard recreation grade receiver with 6 or more satellites in good arrangement and allowed the receiver to run long enough to update the ephemeris, then an error of one minute in lat or long would be very, very unlikely. Maybe a few tenth of seconds of lat/long in bad multipathing situations, but not a whole minute. And although the map is probably great, the person above who expressed unlimited confidence in any map is way off the mark. Flawed maps are put up online all the time. A more likely problem is that the two position sources might be in different datums--there are many, many "flavors" of coordinates out there and this is a very common reason for two sources to disagree. And I would suggest that turning a GPS receiver on and off repeatedly to try to get a good fix is very counterproductive. Leave it running.

 

Good luck!

In my defense I don't believe I said nor implied I had "unlimited confidence" in anything. I said that I would not expect his maps to be wrong to a "significant degree". A minute off is a very significant.

 

Xavier's maps have been used by pro eclipse chasers for years. It is well established by now that he is a careful, reliable source. Infallible? No. But very likely to put out correct info I would trust over most others? Yes.

 

OP needs to do more checking between his GPS, Xavier's map and maybe a secondary GPS or his smartphone. I am guessing it is an issue with his GPS unit or a human error on his part (I make tons of those).



#13 geovermont

geovermont

    Apollo

  • *****
  • Posts: 1,051
  • Joined: 06 May 2016
  • Loc: Vermont

Posted 06 April 2024 - 01:27 PM

Got it. I did not read your original comment carefully. Hopefully do as you suggest and get a second or third position to test against.



#14 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 06 April 2024 - 03:54 PM

I got a reply from Xavier himself:
“ Trust the GPS. The satellite imagery may have an offset.”

Today I went to a few different places to test the GPS. It matches the map coordinates very well. I didn’t have a chance to go back to the observing site yet. I will test it again when I get there. If there is still an offset, I will trust the GPS.

#15 bookemdano

bookemdano

    Explorer 1

  • *****
  • Posts: 58
  • Joined: 05 Jan 2009

Posted 06 April 2024 - 04:19 PM

Ehhh, that was humble of him but a GPS can definitely give erroneous readings for various reasons.

 

Glad to see your latest trials have matched up. Definitely correlate with your smartphone GPS or another handheld if you have one available. There are also other eclipse maps you can compare against, for example Fred Espenak's here: https://eclipsewise....pr08Tgmapx.html



#16 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 06 April 2024 - 04:49 PM

Aren’t all these eclipse maps essentially using Google coordinates? If Google is wrong, all of them are wrong, right?

I also tried to compare the coordinates with my phone reading. All three were different. I use iPhone, and as far as I know, iPhone doesn’t have a real GPS. So I probably won’t trust it unless it’s the inly thing I have.

#17 bookemdano

bookemdano

    Explorer 1

  • *****
  • Posts: 58
  • Joined: 05 Jan 2009

Posted 06 April 2024 - 06:04 PM

Aren’t all these eclipse maps essentially using Google coordinates? If Google is wrong, all of them are wrong, right?

I also tried to compare the coordinates with my phone reading. All three were different. I use iPhone, and as far as I know, iPhone doesn’t have a real GPS. So I probably won’t trust it unless it’s the inly thing I have.

iPhones absolutely do have a real GPS inside. 

 

When you have cellular service, that GPS is assisted by data from the cellular network you are on (known as AGPS IIRC), but even with no cellular signal your iPhone should be able to get your position via GPS satellites. Now is the iPhone the world's best and most sensitive GPS receiver? No. But with a decent view of the sky it should get you pretty close.

 

You say the coordinates were different between the Dual XGPS, your iPhone and the map? How different? With a good fix you shouldn't be more than 10 meters off.



#18 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 06 April 2024 - 06:23 PM

Hi,

 

Unfortunately yesterday I didn't write down the iPhone readings, so I don't know how much is the offset.  My memory is that at that time, the iPhone reading cannot assist me to decide whether I should trust the GPS or the map.  The GPS and the map agreed with each other on latitude, but iPhone did not agree with them.  Then iPhone agreed with one of the GPS and the map (forgot which one) on longitude, while GPS and the map did not agree with each other.  So when I was at the site, I had no idea which one to trust.

 

Hopefully when I go there again, I can get at least two of them completely agreeing with each other.



#19 bookemdano

bookemdano

    Explorer 1

  • *****
  • Posts: 58
  • Joined: 05 Jan 2009

Posted 06 April 2024 - 07:02 PM

When you get on site, power on the Dual XGPS first thing and leave it in clear view of the open sky. Try to leave it stationary for 15-20 minutes (which is more than it needs, but just erring on the side of caution). Then connect SEM to it (or the Dual GPS app) and see what it shows. If it is showing a 3D fix and more than 4-5 satellites then it's probably accurate. You can check it against your phone and Xavier's map. If the iPhone and Xavier's map agree but the Dual XGPS doesn't (at least if it disagrees significantly) then you have a dilemma. You will have to trust one or the other.

 

If they all disagree significantly then just go with the Dual XGPS coordinates and hope for the best. 

 

Hopefully they all agree within 10-20 meters and then you can rest easy.

 

Good luck.


  • whwang likes this

#20 whwang

whwang

    Fly Me to the Moon

  • *****
  • topic starter
  • Posts: 5,185
  • Joined: 20 Mar 2013

Posted 08 April 2024 - 02:01 AM

Not that it matters for other people.  I just want to tell you guys my conclusion.

 

The conclusion is that the Dual XGPS is not to trust in this case.  I let it sit outside for quite some long while (>30 minutes) and read it from the APP.  I did it twice, both getting consistent results.

 

Then I compare the results with iPhone's own GPS reading, my mount's GPS reading, and the map.  The latter three agree with each other, but not with the GPS device.  So I think it's reasonable to just use the map, or iPhone's GPS.

 

As for why this happens?  I have no idea.  XGPS worked well in other places in this city.




CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics