Control Of Google Maps

[Papa-Bear-NYC]

I have on my list a few horizontal control stations that have little intrinsic identification (a nail of a bolt for example). I have been using Google Maps to get a fix on them, but when I get to the location, I swear the thing is not where the photo shows. For example today I found a pike (big nail) and it's the only pike I in the area but the photo on Google Maps shows the location about 25 yards different. Can they be that far off? They show a lot more definition that Topozone, but how good is their control?

 

Thanks -- love to log that pike. Hand held GPS might or might not help.

 

Pb

[Eaglehaslanded]

Well I am not a GIS expert but I do understand some of the problems with mapping programs. When a plane takes a aerial photo, the photo will not be flat. Remember the earth is round. You will not be able to see this. Then there are the different corrs.systems that are use. Photo was taken with WGS 84 and the mapping program LAT/LONG was State Plane.

When you plot your location you could be off a couple of meters. If you look up your address with Goggle Map it could put your house location at the end of your street. This is because the geocode information is off a little.

I don't know if this helps, but someone else might jump in now and give more insight

 

Eaglehaslanded

[+Butano]

This is a pretty cool shot of where two aerial photos were "merged." South San Jose They missed the merge by a bit. As you can see the photo and the map do not line up on the eastern portion of the map. Zoom in a bit and you will see Santa Teresa Blvd make quite an abrupt jump (north west of the pin).

[+Klemmer]

I've looked at Google Map (the on-line one), and Google Earth (the one you have to download). IMHO - Google Earth is much better, and seems to be generally more accurate than my Magellan. Just an impression, not a scientific study. Of course, forget the scaled ones. My data points are from .gpx files processed through BMGPX and GSAK. Quite a valuable tool, IMHO.

[+superpowerdave]

When a plane takes a aerial photo, the photo will not be flat. Remember the earth is round. You will not be able to see this.

 

Just out of curiousity - are these taken by aircraft? Or do they use satelite imagery? Just wondering.

[Z15]

I have noticed this too. I knew the exact location of several BM's on buildings but using the address or postion, google plotted them as much as 200 ft or so off. One was a 1/2 block away, other was across street from actual location.

 

These maps just are not perfect when you zone down.

 

Not Google but another very good map I tried to plot my business. I took actual GPS, entered into the map and it plots a block away at full zoom in but appears right on at a farther out view. I had the webstire manger fool with this for a long time and thats as close as we could get without changes spoofing my acutal coordinates which I did not want to do in case someone used them to find me.

 

In other words, the GPS lat and lon of my shop does not jive with the map. Its close to the second but thats just enough to put it a block off in realilty.

 

I was a survey tech for 30 yrs and the last 10 we used Delorme maps for a lot of recon work, so I know not to trust these computer maps for accuarcy. Some are close in one spot and far off at another. People trust them too much as right on.

Edited April 7, 2006 by Z15

[+Klemmer]

Superpowerdave:

Depends. At "low altitudes" (zoomed in close) they are definitely from aircraft. WAY too good for civilian satellites. At the higher altitudes (like all of CA), they might use some satellite imagery (not sure).

[edmcnierney]

They show a lot more definition that Topozone, but how good is their control?

 

It sounds like you're comparing them to TopoZone's topo maps rather than to our aerial photos (available to subscribers) since our aerial photos (airplanes, not satellites) are much more detailed than Google Maps' for most of the country.

 

If you can give me a description of the location I'd be happy to take a look at TopoZone and Google and give you some comments on the two and what's there. Thanks!

 

Ed

[+GEO*Trailblazer 1]

GOOGLE EARTH PLUS.

I know that everyone I have done so far(BM) is very very close to the mark.

It has much better definition(clarity in imagery) and you can also use your GPSr with it.

 

It is well worth the $20.00 in my opinion.

 

I even have dial up and get it to work fine.

It just takes a while to load.

[+Harry Dolphin]

Since we're talking about benchmarking here, I've come up with some interesting observations.

KV0630 & KV0631

The Topozone map of these. Compare with the coordinate location for these:

KV0630 KV0631.

I think I'll have more luck using the topozone map, for those benchmarks on the topo maps.

 

Then, there's my old nemesis: KV1232 Rockaway.

Notice how the Topozone map actually shows it in the right place (though the NGS coords are wrong? KV1232

That pinpoints the Rockaway Borough hall! I may have to call the borough clerk and ask if I can wander through the building. (The benchmark was/is in a back door stoop. An addition has been added to the back of the building.)

I suspect that when the benchmark shows on a topo map, we're best off using the topo map. We'll see how I make out on KV0630 & KV0631 in the morning.

[+JohnTee]

Had the GC KMZ loaded into GE the other night, looking at a couple of local benchmarks and geocache locations. I know GE is not designed to be highly accurate, but I saw both of those caches jump all over the map - one of them up to 1/10 of a mile!

 

JohnTee

[+Klemmer]

JohnTee: Using the GC to/from GE auto link, the geocache locations are INTENTIONALLY moved around. Part of the deal. Not sure exactly why. It was so noted somewhere in a forum, I think. Probably so as not to make it TOO easy.

 

If you drop a .gpx file onto GE (whether caches or BM's), you won't have that problem.

[+JohnTee]

JohnTee: Using the GC to/from GE auto link, the geocache locations are INTENTIONALLY moved around. Part of the deal. Not sure exactly why. It was so noted somewhere in a forum, I think. Probably so as not to make it TOO easy.

 

If you drop a .gpx file onto GE (whether caches or BM's), you won't have that problem.

 

Well, I DID vaguely remember reading that and KNEW that they moved - just didn't realize they would move up to a tenth of a mile!

 

What I did not know about was dropping a .gpx into GE. Now that's handy. Thanks!

 

JohnTee

[2oldfarts (the rockhounders)]

It sounds like you're comparing them to TopoZone's topo maps rather than to our aerial photos (available to subscribers) since our aerial photos (airplanes, not satellites) are much more detailed than Google Maps' for most of the country.

 

If you can give me a description of the location I'd be happy to take a look at TopoZone and Google and give you some comments on the two and what's there. Thanks!

 

Ed

 

Ed, there is a benchmark at N37°10.226 & W111°57.456. The Photomap for Google is lousy! you can not zoom in enough to see any definition of the terrain. Most of what is available for our area, on Google is of this quality. Are the TopoZone photomaps for this area, any better?

 

Also, can the TopoZone photomaps be downloaded to a laptop and then accessed when NOT connected to the internet? Can GPX files be imported to the TopoZone Photomaps?

 

For us Google maps have little to offer.

 

Thanks for your input.

 

John

[ArtMan]

Also, can the TopoZone photomaps be downloaded to a laptop and then accessed when NOT connected to the internet? Can GPX files be imported to the TopoZone Photomaps?

It's a shame the Google aerial images aren't very good in your area. Google Maps is a terrific tool, as is Google Earth.

 

I don't know about TopoZone photo maps, but you definitely can import .GPX files into USAPhotoMaps, and the aerial images and topo maps are stored locally on your hard drive, so no live Internet connection is needed, and you can take it on the road.

 

-ArtMan-

[mloser]

I love USAPhotoMaps, and use it all the time. It seems very accurate, but I have never counted on it to be 100 percent on, so I haven't put it to the test.

 

Artman is correct that the maps and photos are stored locally, but not the first time you view an area--you have to dowload them, and after that they are stored with the program. Just a warning so you don't think the whole USA is waiting for you in map and photo form, then head out to a mark and find just a blank map. The download process is very fast on a good connection, so getting the info is not an issue.

[ArtMan]

I still don't know whether the coordinates I get from Google Earth are more accurate or less accurate that what I get from my GPSr, or maybe it depends on the quality of GPS signal at a particular moment and the particular patch of Google Earth I'm viewing. (Although, on the handful of benchmarks I've spot-checked, Google Earth has been spot-on.)

 

If you've never used Google Earth, it displays the coordinates in DD MM SS.SS format at the bottom of the screen, but there is no easy way to capture the info. I did discover a nice tool today for capturing coordinate data. It's a small GE file you download, then you center a cursor over your destination, and a pop-up box gives you the coordinates, which you can then copy-and-paste into another application.

 

The file can be downloaded from this site, which also gives instructions for use.

 

One caveat: the coordinates display in DD.DDDD format, and I don't see an obvious way to provide other formats.

 

-ArtMan-

[+Klemmer]

Klemmer's "Back of the Envelope" test:

 

DX4677 (Rose) Adjusted Tri Station, location known to me, and GE has it visually right where it should be.

33 54 56.13434(N) 117 50 33.94884(W) <-- from datasheet

N 33° 54.936 W 117° 50.566 <-- from GC.com page

 

From GE with the new "capture applet":

33.91555 -117.843 cut & pasted

Converted & rounded to Datasheet format (DD MM SS.SS)

 

Lattitude:

GE: 33 54 55.98

NGS: 33 54 56.14

diff: 0.16 second = ~16.2 feet

 

Longitude:

GE: 117 50 34.80

NGS: 117 50 33.95

diff: 0.85 second = ~ 72.4 feet

 

Well, on a sample of one, assuming my BOTE calculation is about right, I say the result is "fair". I normally do better with Delorme Topoquad software, usually about 10 feet, for the same sort of analysis that I have done several times in the past. BUT - GE is a heck of a lot newer than the 1970-ish USGS Topo's!

[+Shorelander]

A few points:

 

You can capture the information relatively roughly without that download - just put a placemark at whatever point you want, right click, go to Edit..., clicked the Advanced check box, and go to the Location tab. Not simple, but it gets the job done.

 

To get more datapoints, you can find relatively short intersection stations, plug in the coords (from either GC or NGS), and see if the crosshairs end up on the object in question. I stress that it should be short, because with your taller intersection stations, since the photo is probably taken from an airplane at an angle, the taller the intersection station, the farther off the top of it will appear to be. A way to get around it is to look at the base and extrapolate the middle from that. NG0540 in Skokie, IL, is a reasonably good example.

 

Finally, as a geographer with a heavy GIS background, I have done some paper-map-to-GIS transitions, and the results can be surprisingly good - or bad. The big problems come at the edges, where a lot of weird warping can take place (at least in my experience). One of the problems you have to deal with aerial stuff is orthorectification, a nice long word that I've never gotten to use in polite conversation until now.

Edited April 25, 2006 by Shorelander

[ArtMan]

Finally, as a geographer with a heavy GIS background, I have done some paper-map-to-GIS transitions, and the results can be surprisingly good - or bad. The big problems come at the edges, where a lot of weird warping can take place (at least in my experience).

 

Interesting point. Although Google tries hard to make it look like Google Earth is one big seamless picture, it is actuall a big mosaic of tiled photos. This is obvious in some places (as when a high-rez photo meets a low rez one), and less obvious in others. But in general, it's hard to know where on a tile you are. Unless there's some trick I haven't unearthed yet.

 

-ArtMan-

 

PS to Shorelander: with your background I hope you will take the time to contribute more to these forums. Your expertise would be most welcome (at least to me).