Map Errors and Pushpin Placement using Google Earth / EPE

 

** Warning** Serious nerd content for those not familiar with map based POI file creation.

I realize this post closely parallels Nats50’s recent “Pushpin Dilemma” thread http://www.poi-factory.com/node/43320. I apologize for any duplication of content.

I’m hoping to continue the above discussion on how to create accurate workable POI files considering the inherent flaws in GE / EPE, GPS position accuracy and user interpretation.

I came across a map error in Google Earth while working with EPE on a POI file in rural WV. It can be viewed by booting EPE or Google Earth and going to: N39 17.192 W080 32.688. The error can be viewed without GE or EPE by going here: http://3planeta.com/evaluate.html and entering the coordinates in the search box. Zoom in until you see the highway divide. Zoom in closer until you see the double image of the power poles. If you were trying to make a waypoint for the pole with the transformer, what coordinates would you use? Where would you put the pushpin? I’m trying to build a POI file for the North Bend Rail Trail which passes through this abnormality. Where do I place the track points?

Small map errors are quite common in GE, especially in mountainous terrain. Most are too small to seriously affect waypoint placement in a POI file and seem to occur at four corner photograph “seam” locations or where there is a change in camera angle. This one however is the most extreme I’ve come across. Relative to GE’s own coordinate system, the latitude appears to be off by .002 degrees (about 12 feet) and the longitude by .016 degrees (about 75 feet). Due to image quality, the problem is not visually evident in earlier imagery but the coordinate error still exists.

Overall, Google Earths position accuracy in North America compared to WGS 84 datum is quite impressive at a reported +/- 30 feet. The actual error varies by elevation and latitude. The compound effect caused by an error such as the one above would probably not be a serious issue when trying to create a waypoint for say a driveway entrance or store location in a file for use with an automotive GPSr. The problem is far more noticeable when trying to create an accurate track or waypoint in a file for use with a hand held unit where there are no terrestrial or map reference points like streets, intersections, signs, etc.

If this isn’t enough, on average, most GPSr’s can reasonably expect 10 meter (33’) estimated position accuracy. Of course, this will vary with location, the number of satellites the unit can “see” and whether or not your unit is WAAS and or GLONAS enabled.

Sometimes these errors will compound and sometimes they will offset each other with no way to predict what will happen. In spite of all this, I’m quite impressed at how well the POI Factory files I’ve used perform in the real world. Still, all these issues should be considered before placing that push pin.

I’m curious how other file creators who use GE / EPE handle these factors and if any have seen map errors of the above magnitude. Do you spot check your files in the field and if so, what are your results? Do you have any tricks, fixes or workarounds to share?

I resemble that remark

ie. the one about "serious nerd"

My answer would be -80.544887,39.286329. because I think that the stitching together of the images results in duplication and I chose the pole I thought was the real one with a transformer.

Now, a question back. You speak of errors compounding. Exactly how does a single waypoint error "compound"?

Can you actually enter W080 32.688 into EPE?

Thanks for the Comeback

jgermann wrote:

ie. the one about "serious nerd"

My answer would be -80.544887,39.286329. because I think that the stitching together of the images results in duplication and I chose the pole I thought was the real one with a transformer.

Now, a question back. You speak of errors compounding. Exactly how does a single waypoint error "compound"?

Can you actually enter W080 32.688 into EPE?

I thought this one would die for sure due its technical nature. I really get into this stuff and get carried away sometimes.

Unfortunately, I’m not great at putting into words what I can clearly see in my mind but I’ll give it a shot.

In an attempt to answer your question, there really is only one error associated with placing a pushpin in Google Earth. The other error occurs when using a GPSr to find that waypoint.

The pushpin error is caused by the inaccuracy of GE itself relative to the standard WGS 84 datum or grid system used by GE as well as the NGS to create their newer topographic maps.

https://support.google.com/earth/answer/148110?hl=en

That error is around 30 feet in most of North America. That means your pushpin sits in the middle of an imaginary 30 foot diameter circle when viewed in GE.

When you go out to find your pushpin waypoint, you and your GPSr are at the center of another imaginary circle. The diameter of that circle is the estimated position error of your GPSr, which is around 10 meters (33 feet) under average conditions.

If these two imaginary circles exactly overlap, the position error of GE is offset by the error of your GPSr. When these circles don’t overlap, the compounded error is the distance between the centers of the two circles.

Remember that when you try to find a waypoint, your GPSr is searching for coordinates relative to the earth’s surface, not a point on a map.

I realize this is putting too fine a point on the process. We all do a pretty good job of placing our pushpins despite these errors.

BTW, I agree with the pole you chose. Now if I can figure a way to place my track across that stitched seam, I’ll be happy.

I don’t really know for sure if you can enter coordinates directly into EPE. I’m far from being an expert with it. I just assumed you could since it’s such a great program. The position errors are caused by GE, not EPE itself.

Understand

bdhsfz6 wrote:

...

In an attempt to answer your question, there really is only one error associated with placing a pushpin in Google Earth. The other error occurs when using a GPSr to find that waypoint.

...

That error is around 30 feet in most of North America. That means your pushpin sits in the middle of an imaginary 30 foot diameter circle when viewed in GE.

When you go out to find your pushpin waypoint, you and your GPSr are at the center of another imaginary circle. The diameter of that circle is the estimated position error of your GPSr, which is around 10 meters (33 feet) under average conditions.

If these two imaginary circles exactly overlap, the position error of GE is offset by the error of your GPSr. When these circles don’t overlap, the compounded error is the distance between the centers of the two circles.
...
I don’t really know for sure if you can enter coordinates directly into EPE. I’m far from being an expert with it. I just assumed you could since it’s such a great program. The position errors are caused by GE, not EPE itself.

You are saying that the error can be "amplified" by the placement of the coordinates that the POI file compiler chooses plus the positioning error of the GPS device. I agree.

However, when we travel to a waypoint - by car or by foot - we are not trying to "blindly" wind up a a specific spot on the ground. We have our eyes to locate the waypoint - so 20 or 30 meters is not really a problem, in my view.

I tested EPE and coordinates are expected to be like -80.545,39.286.

Exactly

The lack of visual reference is the reason accuracy is important in creating POI files for off road use.

Thanks for the tip on the correct format for coordinates in EPE.

Ah!

bdhsfz6 wrote:

The lack of visual reference is the reason accuracy is important in creating POI files for off road use.

...

Thus, the reason for the pole with the transformer.

Had not completely thought thru the off-road experience.

again,

bdhsfz6 wrote:

The lack of visual reference is the reason accuracy is important in creating POI files for off road use.

Thanks for the tip on the correct format for coordinates in EPE.

The majority of the issue isn't about the errors found when using electronic media or devices, it's the false expectation of a level of accuracy that was never promised.

--
Illiterate? Write for free help.

A Very Valid Point.

An error in the waypoint for the nearest Mc Donalds might get you lost but an error in a waypoint for a hiking / biking trail hazard could get you injured or killed.

IMO, GPS users should be aware of the error possibility and make decisions accordingly.