How do you change GPS datum on EoTopo

Submitted: Saturday, Dec 07, 2019 at 16:50
ThreadID: 139400 Views:9702 Replies:9 FollowUps:9
This Thread has been Archived
I have EoTopo on my Samsung S8 and find the GPS position given for a location is out by at least 100 metres.
How can I can correct this? Or is it the phone that requires the datum to be adjusted? What datum does EoTopo use?
Allan
Back Expand Un-Read 0 Moderator

Reply By: Member - wicket - Saturday, Dec 07, 2019 at 18:54

Saturday, Dec 07, 2019 at 18:54
Try:
Settings>
General settings>
Position display format
AnswerID: 629006

Reply By: Dave B18 - Saturday, Dec 07, 2019 at 20:09

Saturday, Dec 07, 2019 at 20:09
Will use WGS84 or GDA94 - wouldn't expect any maps to still be AGD66.
100M is not an issue, and is to be expected in some locations.
Which particular map and map scale do you have this error?
Never been a fan of the Samsung GPS chipset.
Switching OFF A-GPS can in fact give you better accuracy at times.
GPS Test from the Google Play Store is a good free App to test your GPS accuracy.
https://play.google.com/store/apps/details?id=com.android.gpstest
AnswerID: 629009

Reply By: Allan R2 - Saturday, Dec 07, 2019 at 20:11

Saturday, Dec 07, 2019 at 20:11
Thanks for the advice. These settings only change the way degrees minutes seconds is displayed.
AnswerID: 629010

Reply By: Allan R2 - Saturday, Dec 07, 2019 at 20:16

Saturday, Dec 07, 2019 at 20:16
Used three GPS units (Samsung S8, Hema HN7 and an old Garmin). The Hema and Garmin were set to WGS 84, do not know what the S8 datum is. I then went onto Google Earth and got there GPS reading for the same location. The Hema and Garmin where within 5 metres, but the S8 was at least 100 metres out.
AnswerID: 629011

Reply By: The Explorer - Saturday, Dec 07, 2019 at 20:28

Saturday, Dec 07, 2019 at 20:28
Hello - your post is somewhat ambiguous. Do you mean the coordinate readout doesn't match what you expect/know to be the correct coordinates or that the position of the pointer displayed on the screen doesn't match a point on the underlying map?

If you are using EoTopo keep in mind it is a small scale map (most data based on 250K map data from Geoscience Australia) so what you see is not super accurate e.g some roads are shown as lines several hundred metres wide and some are offset to allow for plotting nearby features (e.g rail lines paralleling roads)

I suspect the issue is the map you are using and not the device or settings....but some more info would help establish what your issue is in the first place.

Cheers
Greg
I sent one final shout after him to stick to the track, to which he replied “All right,” That was the last ever seen of Gibson - E Giles 23 April 1874

Lifetime Member
My Profile  My Blog  Send Message
Moderator

AnswerID: 629013

Follow Up By: ExplorOz Team - Michelle - Sunday, Dec 08, 2019 at 16:03

Sunday, Dec 08, 2019 at 16:03
"(most data based on 250K map data from Geoscience Australia)" not true for the last 2 versions of EOTopo, ie. 2018 and 2019 versions. Completely different, new data, highly accurate road alignments etc. Will be accurate to point of both lanes in a road etc.
Michelle Martin
Marketing & Customer Support
I.T. Beyond Pty Ltd / ExplorOz

Lifetime Member
My Profile  My Blog  Send Message
Moderator

2
FollowupID: 903699

Follow Up By: The Explorer - Sunday, Dec 08, 2019 at 18:20

Sunday, Dec 08, 2019 at 18:20
No worries - I was just making an assumption - if EO original sourced data in EoTopo maps is now over 50% then my comment is wrong. I know it is an ongoing project of improvement of original data so ratio of new and improved is increasing on a continuous basis. Certainly better than GA 2008 Map ...not sure what version of EoTopo I have. (PS I was also talking about the Raster version (possibly not relevant in this case now that I think about it) ...not the online/app version. There are unavoidable inaccuracies inherit in a 250K scale raster map).

Cheers
Greg

I sent one final shout after him to stick to the track, to which he replied “All right,” That was the last ever seen of Gibson - E Giles 23 April 1874

Lifetime Member
My Profile  My Blog  Send Message
Moderator

0
FollowupID: 903703

Follow Up By: ExplorOz - David & Michelle - Monday, Dec 09, 2019 at 13:32

Monday, Dec 09, 2019 at 13:32
Greg,

Your comments are correct relevant to the 250K original GoeScience data produced to a raster map sheet however our current dataset are very far removed from the original sources. We have spent many day, weeks, months and years sourcing and in some cases manually rebuilding thousands of roads and positional elements.

But it also should be noted that positional extraction/use of data items from the original GeoScience 2005/2008 data-sets is very much different than looking at or reuse of the Raster product. Even though they had the same name and the Raster was derived/draw from the GeoScience NatMAP 250K dataset, the actual positional data for the POI type features was not out by much provided the Datum's were set/converted correctly. Though also GPS accuracy was not so good back then either so 50/100m differences are not uncommon.

I was puzzled at one point when I went back along a route I researched using OziExplorer and the Raster 2005 setup then went back with 3 devices running ExplorOz Traveller and a much later version of Ozi and found the positional data I had recorded was all out by about 150m however the road lines I had draw were spot on - this was a very strange circumstance and is still a mystery to me, I double checked even the old setup to see what had happened but it does seem like a Dataum problem for the positional entries that did not effect the road lines (why still a puzzle). But having said this all the data has again been updated and is now correct again.

I am sure as we continue to move forward accuracy will get even better and couple this with the massive investments being made in the new Australian Positional Data Framework by GeoScience and the GDA2020 datum - from their Datum Modernisation in Australia projects. We will continue to improve and update our data as part of of major on-going projects in this area.

Regards
David

Now to find out from the OP what the issue he is seeing!
David (DM) & Michelle (MM)
---------------------------------
Currently Mapping in the Field Across Australia Fulltime in 2024 - 2025

Lifetime Member
My Profile  My Blog  My Position  Send Message
Moderator

0
FollowupID: 903717

Follow Up By: The Explorer - Monday, Dec 09, 2019 at 21:28

Monday, Dec 09, 2019 at 21:28
Thanks for info David - EO are doing a great job on that map set.

Just to clarify I was actually just trying to point out to the OP that some maps at some scales are not that accurate - a possible explanation as to why he was not seeing what he expected (without knowing what mapping program or map (i.e. raster or app version of EoTopo) he was using or how he knew what the correct coordinates should be for the POI in question).

I suppose EoTopo (Release 2015.1 raster version is all I have) jumped into my head as an example as I was just out in the desert using it and noticed the location of some water bores and gnamma holes were not that accurate (GA data for sure as doubt you have updated the location of every water bore/rock-hole in Australia originally sourced from GA data).

I will try and provide some coordinate updates on some of the POIs I found errors in near future in addition to the one I have l already sent (after I check online version as obviously the raster EO map I have is out of date :) I have spent way too much money on digital maps and gps units over the years so have yet to jump on the EOTopo app bandwagon - maybe when I retire :)

Cheer
Greg
I sent one final shout after him to stick to the track, to which he replied “All right,” That was the last ever seen of Gibson - E Giles 23 April 1874

Lifetime Member
My Profile  My Blog  Send Message
Moderator

1
FollowupID: 903741

Follow Up By: ExplorOz - David & Michelle - Tuesday, Dec 10, 2019 at 13:29

Tuesday, Dec 10, 2019 at 13:29
Greg, I am sure we can sort out an upgrade download of the Raster 2019 product for you. Touch base with Michelle when you get a moment.
David
David (DM) & Michelle (MM)
---------------------------------
Currently Mapping in the Field Across Australia Fulltime in 2024 - 2025

Lifetime Member
My Profile  My Blog  My Position  Send Message
Moderator

1
FollowupID: 903752

Reply By: ExplorOz Team - Michelle - Sunday, Dec 08, 2019 at 16:12

Sunday, Dec 08, 2019 at 16:12
Hi Allan, are you referring to Places shown on map?

Places data is managed separately to the EOTopo map data and is updateable live.

There are 90,000 POIs in the Places database and some old "Places" are data that was recorded using older technology/less accurate than todays technology but the good news is that our Places database is live so can be updated immediately. And because these Places are used as " Overlays" in the map it means everyone can get updates on demand via the ExplorOz Traveller app so their Places overlays are updated free.

If you can use our system tools please to alert us to each Place that you stumble across that needs a slight shift...we can action within 24 hrs.

You can use the system (app or website) to create a new place at the correct spot and put a comment in the old.
Michelle Martin
Marketing & Customer Support
I.T. Beyond Pty Ltd / ExplorOz

Lifetime Member
My Profile  My Blog  Send Message
Moderator

AnswerID: 629020

Follow Up By: ExplorOz Team - Michelle - Monday, Dec 09, 2019 at 13:44

Monday, Dec 09, 2019 at 13:44
Allan please don't let this opportunity go by - I still haven't received any alerts to shift position of Places you've found to be out by 100m as per your original post. There are a 2 easy ways to do it:

1. on the website, locate the page in Places for the position in question and click the Alert Administrator button. This will capture the Place reference and together with whatever info you can provide us we can make changes.
2. if you own the ExplorOz Traveller app, locate the Place marker for the position in question and click Comments and provide info about the problem so we can fix it. In your case, if you know the position needs adjustment it would be ideal to let us know what the correct position should be - you can either give us the lat/long in your comment, or advise us to delete the original place and then you can use the drop pin feature in the app to create a new Public Place to replace it.

Many thanks for your interest in this issue. Managing a database of 90,000 POIs is a huge and important undertaking and use of public input is of massive importance. We take data accuracy very seriously.
Michelle Martin
Marketing & Customer Support
I.T. Beyond Pty Ltd / ExplorOz

Lifetime Member
My Profile  My Blog  Send Message
Moderator

0
FollowupID: 903718

Reply By: ExplorOz - David & Michelle - Monday, Dec 09, 2019 at 13:56

Monday, Dec 09, 2019 at 13:56
Allan,

I am very interested in your findings and as the app developer and map producer would like to be sure of exactly what we are talking about. Are you out and about looking at things or using Satellite images in the app or something else related to your noted differences.

You mention datum adjustment on EOTopo but are you using the ExplorOz Traveller App or some other mapping program on your device? The Raster map sheets are set to WGS84 via EPSG:4326 so if using OziExplorer or some other app on your phone set the datum to this (though I have configured this default in the OziExplorer .map files we send out with the Raster product). The online maps and ExplorOz Traveller App use EPSG:3857 to align them with the web/online accepted datum Spherical Mercator. This allows us to shift between online map sources without difference. The web site and ExplorOz Traveller App do not give you a datum adjustment option as they are locked to the production datum.

As mentioned by several people certain devices behave in different ways and AGPS/GPS/GLOSNASS, power management, cell tower triangulation, wifi detection and a wealth of other technologies can mess things up. But it is important to me that we identify any possible data errors to get a plan of action to check/remedy any potential issues. Can you point me at some of the places you have noted and is it the places layer shown on the app or was it a feature drawn on the underlying map data that was different.

As mentioned by Michelle above there are two sets of data at play in the app, one being the base map EOTopo and the other POI dataset being Places shown with the markers. These two layers are melded together during EOTopo updates so it is vital to us that we are sure we have the best Places data layer before we start on the next round of EOTopo updates though we do not use all Places data in the map production as a great number of Places elements are not needed on the Topo product.

Anyway if you can let us know we can work with you to identify what is going on, best to connect with us using the Help Desk as this allows us to manage it and track the issue to completion.

David
David (DM) & Michelle (MM)
---------------------------------
Currently Mapping in the Field Across Australia Fulltime in 2024 - 2025

Lifetime Member
My Profile  My Blog  My Position  Send Message
Moderator

AnswerID: 629029

Reply By: Allan R2 - Monday, Dec 09, 2019 at 22:22

Monday, Dec 09, 2019 at 22:22
Hi David and Michelle
I am not referring to Places on EoTopo.
I did a GPS test with the three GPS units (Hema HN7, an old Garmin both set to WGS84 datum and Samsung S8 (do not know what datum it uses) with the latest EoTopo maps. I used the middle of our 10 hectare property in the hills and took a reading from all three GPS units. This position is easily visible on Google Earth. The Hema and Garmin were within 5 metres, but the Samsung S8 with EoTopo app gave a different reading which is about 100 metres to the SW.
My wife and I are bird watches and all of our Australian bird sightings are entered into eBird. Within this program you enter your bird sightings and GPS coordinates which then takes you to a map that can be either satellite or what looks like Google maps. This is we have found the Samsung S8 out. When travelling we use the car based Hema HN7 GPS readings for our locations. But when we walk away from the vehicle the offline EoTopo maps are used, but give us incorrect location readings.
The problem may well be the Samsung S8 having a different datum, but I do not know what datum they use.
Hope this helps
AnswerID: 629045

Follow Up By: Member - Warren H - Tuesday, Dec 10, 2019 at 12:02

Tuesday, Dec 10, 2019 at 12:02
Being a retired person currently hibernating inside away from the heat and smoke I took to the net and came up with a couple of interesting links. The possible truncation/rounding of coordinates when passing from the gps to the application seems to be a possible explanation? Links below.
I have a passing interest in datums and coordinate transformations, as from time to time, I do a little work taking historical data sets with geolocated observations, often from mapsheets/airphotos and recorded in a variety of datums such as AGD66/AMG 66, AGD84/AMG84 and recorded as either geographic or grid coordinates and reformat the data for import into a national database.

1. Without bending one’s brain with the statistics and skimming the intro, discussion and conclusions this is an interesting read:Comparative analysis of positioning accuracy of Samsung Galaxy smartphones in stationary measurements.

2.Galaxy S8 Photos : GPS coordinate rounded in Exif..

NT Pajero
2007 Goldstream Crown

Member
My Profile  My Position  Send Message

0
FollowupID: 903750

Follow Up By: ExplorOz - David & Michelle - Tuesday, Dec 10, 2019 at 13:25

Tuesday, Dec 10, 2019 at 13:25
Allan,

Thanks for the info, if you are simply looking at the GPS position displayed on the screen it is displayed exactly as reported by the GPS unit in the device - there is no transformations or data management at all. Are you using DD, DM or DMS for the position format? The GPS reports DD and I have some formulas that convert to DM and DMS so there maybe something for me to look at there if you are using these, but I am using floating point data types for the maths routines but there is some rounding going on.

As mentioned I am very keen to ensure our products are not causing any issue, if you can let me know the position format you are using in Settings? And also if you are just taking the Lat/Lng data from the screen as your positional reference?

David
David (DM) & Michelle (MM)
---------------------------------
Currently Mapping in the Field Across Australia Fulltime in 2024 - 2025

Lifetime Member
My Profile  My Blog  My Position  Send Message
Moderator

0
FollowupID: 903751

Reply By: Allan R2 - Tuesday, Dec 10, 2019 at 21:56

Tuesday, Dec 10, 2019 at 21:56
Michelle
I am using DMS format.
I have just taken a GPS reading with the Hema HN7 and EoTopo on the Samsung S8. There is only 0.4 of a second difference between the two. Thats very good. Don't know what the issue was the other day.
Apologies for wasting your time.
Thanks to you and your team for their thoroughness in trying to solve my problem. And. To others that came in on the conversation.
Cheers
Allan
AnswerID: 629052

Follow Up By: The Explorer - Wednesday, Dec 11, 2019 at 00:34

Wednesday, Dec 11, 2019 at 00:34
Hi...

I know what the issue was in this case..had it several times today. GPS signals/reception are not always perfect under certain circumstances and you need to be patient or just accept the errors.

Also for bird records, I wouldn't be phased by a 100m error in any case, they tend to move :). A 100m ÷/- location for a sighting is actually very good.

Cheers
Greg

I sent one final shout after him to stick to the track, to which he replied “All right,” That was the last ever seen of Gibson - E Giles 23 April 1874

Lifetime Member
My Profile  My Blog  Send Message
Moderator

0
FollowupID: 903770

Sponsored Links