Is it possible to use Google Earth/Maps to know how big a track is???

Aghast wrote:

See:

http://www.gmap-pedometer.com/phpBB2/viewtopic.php?t=12211

"Big ups to our new elevation provider,

http://www.geonames.org/

, and NASA! (The elevation data was collected on a mission of space shuttle Endeavor in 2000.) "

AND

http://www.gmap-pedometer.com/phpBB2/viewtopic.php?t=3

"API mapping data from...Teleatlas"

I could see that one database gives the elevation of, lets say 120 meters at a stadium, while another database says it is 133 meters, but how does that give elevations changes

on known perfectly flat areas like tracks? Whether you use Database A or Database B shouldn't matter, Tracks are flat.

If you want to see an example of multiple databse issuess go here.

look at the elevation data of any marathon on the front page then click on the map link. (example Boston)

Then check the elevation box in the "info and Tools" dialog box. Compare the data.

I've used google maps to plot some of my running routes and the elevation data shows that I ran down a hill into a deep valley of 100m below the hill and then back up another hill.

In actual fact I was running along a ridge and the elevation change would have been less than 20 metres.

malmo wrote:

No disagreement with the first part. GPS are still toys on par with gamboys and Ps2.

I think Malmo is splitting hairs. Running with a GPS unit is known to be inaccurate because of sampling and connection errors -- I think this is what he's talking about. GPS itself is extraordinarily accurate, although of course it depends what precision you're hoping for. Gmaps pedometer is quite good as well.

There are some small absolute errors. For example, the road doesn't always match the satellite image, so there is some small error there. But certainly not 0.18 miles instead of 0.25 -- maybe a couple of meters on an oval track. If you ride your bike and plot the GPS route in garmin connect, it lies right on top of the road (not the satellite image) every single time. I'm looking at some plots right now and I could only find one data point over a 20 mile ride that doesn't sit right on top of the road, and I think it was because there's a large shoulder on a 90-degree turn. Sure, there's some sampling error so the distance might be underestimated, but you shouldn't confuse sampling error with device precision.

GPS/Google Earth - toys?

Sorry but I strongly disagree.

As one of the founding members of the "ultra nerd group of measurers" I found them both to be a valuable and time saving tool....not a toy.

Google Earth

I can create a course using Google Earth in minutes what use to take as much as several days. In the past, I would go out and measure several segments of roads trying to fabricate race courses but now it can be done with relative accuarcy with Google Earth. When done carefully, I can get within 10-20 meters accuarcy of a 10k course using GE. Of course I would still go out after and do the two bike (jones counter) measurements for certfication.

GPS

Just yesterday I measured a 5k course in a state park on paved roads that had very few landmarks. I used PK nails to identify the start, finish, and mile marks. I also noted the lattitute and longitude of each of those points with my handheld GPS.

I went back several times to those points to see if the GPS would identify those locations (to find the Pk nails). Each time the GPS was within 5 feet.

Tools yes, toys...not how I use them.

Now now David. Some of us ultra-nerds have looked at huge mainframes with rotating iron pipes for storage as toys at various times in our career, and yes, it's how you're using them at the time.

Hell, I was one of the first measurers using computer generated maps (long before MapQuest or Google earth) that submitted paperwork via email as PDFs (after I got my state certifier to step away from the FAX machine and physical mail). Hell, I put Garmin GPS locations on maps 10 years ago. Back when people would look at them and ask "what's that"? A big honkin' Garmin, and you had to get your map data on CD. I have the forms loaded as Excel files I use on a PDA I take with me while measuring so they auto update and they're ready to print when I'm back.

So you want to see who can spray their nerd-dom higher on the wall, I'm game.

GPS, Google Earth, and all the other fun stuff we use can be tools in the hands of people that understand their proper use and limitations. Just like a hammer in the hands of a 4 year old is a toy, one in the hand of a finish carpenter is a tool. Just because Picasso could do some amazing s**t with a paintbrush doesn't mean a nice paintbrush is all I need to be an artist.

This is my main beef with this line of discussion I see crop up. Sure, GPS and other tools CAN be accurate under many circumstances and to counter Malmo's line, are probably close enough for this particular government as far as the OP is concerned. Hell, lying about your training times goes back to Phidippides so the OP might as use it as an opportunity to pad his a tad too. It's a round oval, must be 1/4 mile.

Just last night at a Christmas party I had a guy (first time half marathoner) tell me he was disappointed in his time, but then he found out from a guy wearing a Garmin that the course was long so he felt better about it. I measured that f*&king course, probably 4-5 times over the last 15 years with the various road changes.

Prior to that same race three other guys I know went over the course, one on a bike to put down some spotter marks to help find the mile markers, two others running wearing the same Garmin model. It wasn't intended as an experiment but the end result was the guys came away with "wow, if I run the tangents these Garmins are accurate". Two guys, same Garmin, same course, same day. 13.10, 13.13. What some see as a triumph of accuracy I see as two measurements being off by 150 feet. I'm picky like that.

So yes David, in the hands of 99.8% percent of the population I look at Google Earth, GPS, and all other gadgets as a toy. I personally use them every measurement as I have used these types of things as tools for years, but I would never say that they are accurate in the hands of the unwashed masses and if you haven't gathered by now, it kinda pisses me off when other people do.

Keith Stone wrote:

Hell, I was one of the first measurers using computer generated maps (long before MapQuest or Google earth) that submitted paperwork via email as PDFs (after I got my state certifier to step away from the FAX machine and physical mail). Hell, I put Garmin GPS locations on maps 10 years ago. Back when people would look at them and ask "what's that"? A big honkin' Garmin, and you had to get your map data on CD. I have the forms loaded as Excel files I use on a PDA I take with me while measuring so they auto update and they're ready to print when I'm back.

Keith, do you have any idea what the acutal Garmin algorithms are?

A while back someone pointing out the published accuracy from the Garmin Manual. (Don't quote me) It was on the order of 4-5 meters as I recall. My understanding was that meant any given fixed point was going to be accurate within 4-5 meters. (presumably 68 % of the time).

SO does that mean that the GPS device measure the cumulative sum of the incremental measurements (links) meaning that the accuracy problems (assuming a straignt course) with GPS devices are off by simple trignometric differences from the straight line?

Or does the algorithm assume a normal distribution of data points, then 'recenters' them, then calculates the summation of the individual links?

Am I making any sense to you?

What is the frequency of data samplings fro a GPS device? How many times a minute?

For the new online running log website I am making, I've also made a course measuring tool (pretty nice if I say so myself), and I use the Google Maps API for both distance and elevation.

Malmo is correct. While the distances can sometimes be accurate, there is no guarantee. These are simply satellite photos which are stretched and resized to roughly fit lat/lon coordinates. For training purposes, it will give you a good idea of how far you ran (I'd say usually accurate to a 1/10th of a mile), however, for racing purposes, Google Maps should not be used at all. It just isn't that accurate. I believe somewhere in the API, Google even states this (too lazy to find the link).

For direct distance calculation between two points, I use Vincenty's formula (http://en.wikipedia.org/wiki/Vincenty's_formulae#Inverse_method) which is accurate to 0.5 mm. A lot of course mappers use this. But here's the problem: this formula doesn't take into account the changing elevation and only measures a direct route on the earth at sea level. Furthermore, as I stated before, Google Maps cannot give you a precise latitude and longitude for a location because the images are only roughly stretched to fit the coordinates.

Elevation is even more inaccurate than distance. Google obtains altitude data for certain locations on the earth, and then interpolates everything else. These should only be used to roughly gauge how hilly the terrain is.

Keith,

Refreshing comments from someone who actually gets their hands dirty doing this stuff.

Have a great holiday. See you on the calibration course!

David

Keith Stone wrote:

So yes David, in the hands of 99.8% percent of the population I look at Google Earth, GPS, and all other gadgets as a toy.

The same could be said of the Jones counter, no? Any measurement device in the hands of someone who doesn't know how to use it is a "toy."

But that's really beside the point. The OP wanted to know if Google Earth can be used to measure a track. The answer is yes, it can be used to measure a track to an accuracy of about 2-3 meters if you are careful.

malmo wrote:Keith, do you have any idea what the acutal Garmin algorithms are?

Malmo I'll admit I've made a number of assumptions based on my research of what they are. What I've read about within 5 meters works very well sometimes and not others. I see the collie effect periodically when in theory it should not happen.

Back when the military dropped their objections to the accuracy of civilian GPS I was thrilled to death. A lot of what I would do with GPS would work great. A significant percentage would not for a reason I could not explain. I find that troubling. I find the same thing with the various online mapping tools.

So as far as that "am I making any sense to you", sure you are. I'd sure as s**t like to find a sure fire explanation. Maybe I'm too anal but when I go over the same route multiple times with the same tool and get different results it bothers me. As far as what goes in my log well who gives a crap. That's why I say if you're measuring a training run I don't give a rat's ass if you're measuring it with a 1968 VW Microbus. The gal in the typing pool you're trying to boink doesn't know the difference anyway.

dukerdog wrote:The same could be said of the Jones counter, no? Any measurement device in the hands of someone who doesn't know how to use it is a "toy."

But that's really beside the point. The OP wanted to know if Google Earth can be used to measure a track. The answer is yes, it can be used to measure a track to an accuracy of about 2-3 meters if you are careful.

No doubt, and I'm not arguing that point.

However, on a track with potentially sharp circumferences the zoom level available in that location could alter the measurement significantly. You also have to admit it's possible to "measure" an oval track with 6 or 8 points or reference and the frequency of those will significantly affect the distance measured when we're talking about a curve. Correct?

Again, between you, me, and the guy next to both of us in a bar who gives a s**t. Between you, me and and the guy that's signing the paperwork to be reviewed by a Reigel maybe there's a higher bar.

You yourself said "if you are careful". It's quoted right above this. We're both anal and we'd put 30 or more points around the track. So when the question is "can you use x tool to get an accurate measurement" the answer isn't 'yes' the answer is 'yes if you can do this', and even better if 'yes if you can do this twice and get the same answer'.

So if we really go back to the OP's question you run a hard lap once and write down how long it takes. Rest 10% of that and repeat until you barf. Problem solved. You want to set a record on it, call us back.

I have a friend who is a physicist at Johns Hopkins and does oceanographic experiments. He mentioned that GPS technology could detect a mid-ocean tsunami (about one foot height) with GPS satellites. I said that I thought that was amazing. I said, you mean you can triangulate the position of a spot on the earth's surface that is rotating at 1000 miles per hour and is not a perfect sphere, with three satellites orbiting the earth in different directions at different altitudes and at different speeds (approx. 8000-10000 miles per hour), and the surface of the ocean is being pulled by the position of the moon AND the winds AND the depth and shape of the ocean floor and coastline AND the temperature AND salinity of the water? With all of these variables, you can identify a change in depth of the ocean as little as one foot? That's amazing! How do they do that?

His response: "good question."

No offense but you sound like a real jerk to your friend. Did you ever think that perhaps you don't know everything?Maybe this is what he was talking about.http://www.techbriefs.com/component/content/article/6342"GPS stations can detect ground motions of earthquakes in real time, as frequently as every few seconds. In the present method, the epicenter of an earthquake is located by use of data from seismometers, then data from coastal GPS stations near the epicenter are used to infer sea-floor displacements that precede a tsunami. The displacement data are used in conjunction with local topographical data and an advanced theory to quantify the destructive potential of a tsunami on a new “tsunami scale,” based on the GPSderived tsunami energy, much like the Richter Scale used for earthquakes. An important element of the derivation of the advanced theory was recognition that horizontal sea-floor motions contribute much more to generation of tsunamis than previously believed. The method produces a reliable estimate of the destructive potential of a tsunami within minutes — typically, well before the tsunami reaches coastal areas.The viability of the method was demonstrated in computational tests in which the method yielded accurate representations of three historical tsunamis for which well-documented groundmotion measurements were available. Development of a global tsunami-warning system utilizing an expanded network of coastal GPS stations was under consideration at the time of reporting the information for this article. "Or thishttp://www.google.com/url?sa=t&source=web&cd=3&ved=0CCcQFjAC&url=http%3A%2F%2Fwww.terrapub.co.jp%2Fjournals%2FEPS%2Fpdf%2F5210%2F52100841.pdf&rct=j&q=gps%20detect%20tsunami&ei=YHAVTeTLD4et8Aa017GbDg&usg=AFQjCNHS965FYR4llfHpOH_8E5A5I_n_wA&sig2=FRpZizF6tVmcmvGmhcmpqA&cad=rjaOrhttp://www.google.com/url?sa=t&source=web&cd=4&ved=0CDIQFjAD&url=http%3A%2F%2Fbowie.gsfc.nasa.gov%2F697%2Fstaff%2Fcolombo%2Fpdfs%2Ftsunamis.ion.ntm2000.pdf&rct=j&q=gps%20detect%20tsunami&ei=YHAVTeTLD4et8Aa017GbDg&usg=AFQjCNEKL4DOj79-BrSdDoGZWROt-8reTg&sig2=2rB8jwbxiX4yr-F5B3Herw&cad=rja

Aghast wrote:

No offense but you sound like a real jerk to your friend. Did you ever think that perhaps you don't know everything?

Sorry pal, your intent was to offend. Mine was to further discussion. And no, your link was not what we were discussing. They calculate mid-ocean depth changes with bouy/transmitters that are anchored to the ocean floor.

Anchored to the ocean floor but their relative vertical displacement is measured by GPS (akin to a garmin GPS watch). At least that is what the second link refers to. That is why it is germane. It is accurate enough to detect a tsunami.

malmo wrote:

Aghast wrote:

No offense but you sound like a real jerk to your friend. Did you ever think that perhaps you don't know everything?

Sorry pal, your intent was to offend. Mine was to further discussion. And no, your link was not what we were discussing. They calculate mid-ocean depth changes with bouy/transmitters that are anchored to the ocean floor.

Yes, actually the 2nd and 3rd links went directly to his work. My point remains - amazing.

dukerdog wrote:

The same could be said of the Jones counter, no? Any measurement device in the hands of someone who doesn't know how to use it is a "toy."

But that's really beside the point. The OP wanted to know if Google Earth can be used to measure a track. The answer is yes, it can be used to measure a track to an accuracy of about 2-3 meters if you are careful.

Interesting. I wonder if Google Earth can come within 2-3 meters of a measured distance of 400 meters if you didn't know beforehand what the distance was?

Heck, early in this thread "Aghast" said he measured his track using gmaps at 400 meters. Problem is tracks are not 400 meters around they are 398 meters. Had he know that beforehand I'm sure he would have tweaked is sampling points to come up with 398 meters.

Then "unholy Cumulus" measures his track at 398.8 meters. My question is: is there anyway to mask the running total on gmaps so that you could only make 'blind' measurements? If so, how close can you get? Multiple trials are fine with me. What is the real accuracy of using gmpas to measure a known measured course?

malmo wrote:

My question is: is there anyway to mask the running total on gmaps so that you could only make 'blind' measurements?

Maybe not with the gmaps interface directly, but there are other websites out there that will let you play around with points in Google Earth (without displaying a total) and then dump them out to a text file. These could be put in to Excel (or similar) and then the gmaps distance formula could be applied. Gmaps used to have the formula they use posted somewhere on their website.

Well since all the big races are run by a government, Government work devices used for measuring are good enough.

GMAPS has become more accurate over the years.