Differences in Google Earth and Garmin Mapping

Just curious. But, why would it matter?

I did a similar thing for a train ride through the Rockies. Fun, but not really critical to anything.

IMHO it's not worth worring about. The train is not going to go off track because your GPS route is 30 meters off.

It would be interesting to see if any other on-line maps are more accurate than Google Earth. I have been getting some experience with OpenStreetMaps, MyRouteApp and Garmin Explore.

Also, when I did some railroad POIs years ago, I found a source of pretty precise data that showed the actual GPS coordinates of railroad crossings. I can't recall right now if it came from the FRA or somewhere else, but it seemed to be highly accurate.

There are very slight inaccuracies somewhat dependent on the system being employed. In that respect you might want to look at the coordinate conversion site at https://coordinates-converter.com/en/decimal/40.748428,-73.9.... Note: I've pre set up for the Empire State Building. You can readily change the site by using the Search field at the top right of the screen. The search works reasonably well on my Win10 machine; so-so on my iPad. A bit "buggy" with USA locations, perhaps since the site is from Europe.

Anyway you can see the coordinates based on about nine different systems in use, even including the "What three words" system that defines the Empire State Building as "march.dragon.shield." That system by the way maps the entire earth based on a 3 x 3 meter block, every block having its unique three word descriptor.

I am sorry I may have not made my application clear. I volunteer with the National Parks Service on Amtrak. I provide interpretation on 250 miles of the Empire Builder. With a train traveling at 80 MPH or 117 feet per second, interesting views on the train pass out of view sometimes in seconds.

So I programmed an old Nuvi 2555 with POI and a radius to alert providing a warning of an interesting POI approaching.

I am an old hand at garmins and POI's and a fixed guideway mode of transportation should be a no brainer for the old Nuvi. It failed miserably. So I started to review the POI locations. On google Earth the lat/long are dead center on the tracks. When you look at the POI on the Nuvi (the 'I" button, the POI can be as much as 30+ meters off the tracks.

So posing the question again, does the error in mapping matter? Is that the cause of the miserable performance of the Nuvi?

There is a user-friendly coordinate converter available to you at:

https://www.ngs.noaa.gov/NCAT/

I just tried my house in Albuquerque NM, asking the difference between NAD27 (which is what all my old good maps use) and WGS84 which is generally what GPS systems use until told to set another datum.

Here the stated discrepancy was about 5 meters in latitude and 15 in longitude.

It will depend on location, and which particular schemes you compare.

Using on a train in interesting terrain, you may also have gotten receiver error coming from multipath problems, with the receiver using at least one satellite signal arriving via a reflection (either inside the train, or from nearby terrain). That would jump around. A datum related problem would change quite slowly with position.

You might like looking at the FRA Safety Map.

You may click on mileposts and they will display the GPS coordinates.

Of course, if the map is wrong, this might not help very much!

https://fragis.fra.dot.gov/GISFRASafety/

Don't forget that the Nuvi is a "car" GPS, and the maps it uses are mainly for roads (a lot of dirt roads are not even mapped). I have used my Nuvi's on off-road trails, especially around Silverton and Durango, CO, but they are not as accurate (within reason) as they are on paved roads.

If your POI location can be up to 30+ meters off, but you're traveling at 117 feet per second, then it seems the difference between the actual and indicated POI location will be about one second. If you need to know of an upcoming POI so you can let the passengers know to look for it, won't you need 10-15 seconds to make the announcement and give them time to look in the right direction?

When you say you're setting a radius alert, does that mean you're using Garmin's TourGuide? I'd think you'd want a good 1000+ foot alert to be set, given the purpose of your POI file. If not using TourGuides, I'd think that some of the POI locations may not offer an alert. Another issue could be Garmin's "snap-to-roadway" feature that may inadvertently show your location along a nearby road rather than the rail line.

If it does turn out that Garmin's map shows the wrong location for the rail line, that's going to be caused by the map maker, Here, not Garmin.

Just a couple of thoughts here, from a past White Pass & Yukon Route tour guide. Ah, the stories I could tell

This may or may not answer your question but here are a few thoughts.

I do volunteer work for the USCGS by recovering benchmarks and am very familiar with this type of error.

Just because the lat/lon "pin" you drop in Google Earth is dead center on the tracks is no guarantee of accuracy. Although GE is fairly accurate 90% of the time, there are inherent errors that can show in a variety of ways. One, for example, can be seen by adjusting the time slider in GE and seeing if the pin moves in relation to the image.

As already discussed, there are also errors in the Here map database.

Your older Nuvi2555 isn't as sensitive as newer GPSr's, especially those that are GLONAS compatible. The metal shell of the train car with it's lighting & wiring act as a shield which can further attenuate the satellite signal. Check the EPE, or estimated position error, on your GPS. you might find it to be close to the 30M error you're seeing.

Position errors can sometimes compound or offset each other which further adds to the confusion.

As posted in the Subject line, is it definitely the map and not the GPS receiver? The reason I ask that is because GPS technology has evolved since I have been following it the last 20 or so years. For example, the government used Selectively Availability until May 2000. Selective Availability (SA) was an intentional degradation of public GPS signals implemented for national security reasons.

This is a summary of things regarding GPS accuracy.

The GPS: The Global Positioning System site has a good description of accuracy. It will depend on the quality of your equipment, how many satellites are visible and processing in your equipment and the quality of the GPS signals (direct, reflection, interference). As measured by the Air Force that runs the GPS program, the error across the entire constellation of satellites was 2.1 feet, 95% of the time on 20 April 2021.

There are a lot of factors that affect individual performance at any time. If you measure the accuracy of your unit in an open field or on the ocean you could get different but probably close numbers every day. The weather affects the signals. Also the satellites have to know exactly where they are in 3 dimensions. If an orbital change has occurred, using that satellite will yield a wider error ellipse but probably still pretty close.

In other words, buying better equipment does improve the accuracy, but do you care? Driving a car, if I am off by 5 or 10 feet, it is a don’t care. Accuracy to less than an inch only maters with surveyors and they use fixed references on the ground to measure those inches, not GPS. Better GPS however does give you a very accurate time mark. So if you want to know precise time or have a need for that, you can set your standards to GPS. In most applications it is not the exact time as much as it is all of your stations working from exactly the same time base.

I still recall some posts about the SiRFstar III chipset. What GPS chipset does your Garmin 2555 use?

SiRFstar III is a GPS processing chip manufactured by SiRF Technology, Inc. It is used in some newer portable GPS units such as the Garmin GPSMap 60CSx. It is able to acquire a fix much better than older chips. It works in many cases in narrow streets, in trains, sometimes even indoors, where older chips don't get a useable signal any more. The chip uses 200,000 correlators to fix the position, and will operate on 20 channels with signals as low as -159 dBm. Also supports WAAS.

https://www.gps.gov/systems/gps/modernization/sa/

Auto GPS units have a "snap to road" algorithm which positions the display point (and I assume any POI) to the nearest road. I believe it moves about 30 meters but can be mistaken.

I don't think lkoehnen's issue is with the GPS current location, the location of whatever icon was chosen to be the vehicle. I do think the OP's concern is the designated position of a custom POI location as detailed in a POI file, and where it's shown on the Garmin display with City Navigator NA, then compared to seeing the same POI location on Google Earth.

lkoehnen, what do you see on Google Maps, not Google Earth? Does GMaps match GEarth or is there the same issue as with Garmin's map from Here? Just curious.

My Garmin Zumo XT has a Usage Mode and Off-road mode might disable that, but I don't know that Garmin Nuvis have that feature.