Gps Sped Accuracy On Hills?

[+wickedsprint]

Since gps speed is a form of ground speed, if one were ona steep enough hill wouldn't it show the speed as lower? Simple geometry..the hypotonuse is longer than the direct ground line.

[+Jamie Z]

if one were ona steep enough hill wouldn't it show the speed as lower?

Yes, but the simple part is that even on the steepest hill you could find, the difference would be almost negligible.

 

Jamie

[+KaiserKlan]

I was thinking about this very thing. We took a trip of just over 1200 miles, I had reset my trip indicator on the GPSMap 76s to zero and my car. When we got back the car showed 7 more miles than the GPS, the only thing I could figure was the hills in-between. Possible?

[+WeightMan]

Just a note: seven miles is only 0.58 percent. Not a bad error rate.

 

Car odometers are not the most accurate things in the world as was pointed out in another recent thread. I think that is more a tribute to the odometer than anything else. As tires wear the diameter (and circumference) becomes smaller and the number of rotations per mile increases. With old tires the odometer will read long since all it does really is read rotations of the tires.

[+Sputnik 57]

I was thinking about this very thing. We took a trip of just over 1200 miles, I had reset my trip indicator on the GPSMap 76s to zero and my car. When we got back the car showed 7 more miles than the GPS, the only thing I could figure was the hills in-between. Possible?

More likely that your car odometer is slightly off. 7 miles out of 1200 is pretty darn close.

[+Absolut&KittyKat]

Hope this makes sense...

 

Your cars odometer is based on how many times your wheels rotate.

 

If you have a 195/60R14 tire, your tire diameter is 23.2" (tirerack.com)

Pi x 23.2" = 72.9" circumference.

Thats 869 revolutions of the tire per mile. (23.2" into 5280ft)

 

Now, most tires start with 10/32" of tread depth, if your tires are worn lets say 1/2 way down, your tires are smaller by 10/32" (you wore 5/32" off each side of the tire"

 

10/32" = 5/16" = .3125"

Original tire diameter = 23.2" - .3125" = 22.88" = 71.9" circumference

Now your up to 881 revolutions per mile.

 

So even though you went the same distance, with worn tires having a smaller overall diameter theyhave to rotate more to cover the same distance. When your tires rotate more, the odometer rotates more. ie: showing more miles.

 

This inaccuracy is compounded by underinflated tires.

Not to mention that most speedometers are inaccurate to begin with.

 

Also, you may not have had gps lock the entire time you were moving, plus that fact the gps only measures distance covered from point to point not changes in altitude of the road. Your original idea of hills is correct, albeit a small differance, over a very long trip with many hills it would affect the overall milage by quite a bit.

[+Absolut&KittyKat]

Car odometers are not the most accurate things in the world as was pointed out in another recent thread. I think that is more a tribute to the odometer than anything else. As tires wear the diameter (and circumference) becomes smaller and the number of rotations per mile increases. With old tires the odometer will read long since all it does really is read rotations of the tires.

Thats the easy explainition

[GPS togo]

Doing some quick calculation on the GPS (horizontal) speed vs. the true down hill speed, I found it to be somewhat interesting. And thought I would pass it along to anyone who may be interest in it.

 

If my trigonometry and vector analysis is correct (it has been a long time), to get your true ground speed going down a slope, you would divide the horizontal (GPS) speed by cosine A (A is the angle measure between the horizontal line and the downward line of the slope). Or stated another way: true slope speed = GPS speed / cos A.

 

Another way of looking at this, is that cos A x 100 will give you what the percentage of your GPS speed is of your true slope speed.

 

For example, take the following three slope angles:

 

10 degrees (probably a moderately steep road grade)

30 degrees

45 degrees (probably a moderately steep ski slope grade)

 

cos 10 = .984 or 98.4 % of your true slope speed

cos 30 = .866 or 86.6 % of your true slope speed

cos 45 = .707 or 70.7 % of your true slope speed

 

At 50 mph GPS speed the differences are:

 

10 slope -- true slope speed is 50.8 mph

30 slope -- true slope speed is 57.7 mph

45 slope -- true slope speed is 70.7 mph

[+Jamie Z]

10 degrees (probably a moderately steep road grade)

45 degrees (probably a moderately steep ski slope grade)

I didn't go through your math (I just got out of Calc III about 30 mins ago, so give me a break ) but these statement are not true.

 

10 degrees is much more than a moderate grade road. I don't have my calculator or the time handy right now, but a 10% slope is somewhere around 20% grade, which is far more than you'll see on any regularly traveled road.

 

Ski slopes are usually in the area of 10-20 degree slope, and although there may be a handful in the world, only the most elite athletes would ski a 45 degree slope.

 

See my post here and the accompanying discussion for more and better information.

 

I said a year ago, "The problem is perception. Humans have a tendency to greatly overestimate gradient. It doesn't take much hill to seem steep."

 

Apparently I've taken on the lifetime goal of educating people that slopes have almost no effect on GPS indicated speed.

 

Jamie

Edited September 15, 2004 by Jamie Z

[+WeightMan]

On freeways slopes greater than about 4% are marked as such for truckers. The highest I have seen is 6% which would be 3 degrees. Without doing the math the difference would be minimal given than 10% yields 98.4%. Now mountain roads may get steeper slopes for short distances. Since mountain roads around my area are built for log trucks the slopes won't be that great.

[Wild T2]

I seem to recall that the Interstate Highway System was to have a maximum grade of only 4% (4 foot rise or fall in 100 feet). But there were places where that was not economically feasible, so it didn't happen.

 

I don't have my references handy, but I think that a 9% grade is the maximum for a highway that receives Federal funding. I bet that there is a process to get a variance from this requirement, but I suspect that it is a very difficult process.

 

The steepest main-line railroad grade in the country is in otherwise flat Indiana. Its just under 6%.

http://www.oldmadison.com/madview2.html

[+Pen&Card]

I'm glad this came up. Rather than changes in elevation, I've always assumed there had to be some margin of error - always on the short side of measuring speed and distance, when the GPSr is tracking a curve.

 

Although my GPSr reads once per second, there is no possible way the GPSr can measure the true length of distance traveled in the arc between readings, correct? At hight speeds, the curve would have been sampled in a few straight lines, and possibly significantly shorted the GPSr measurement of a 1000 mile or more trip?

 

It would seem to me this would create more inaccuracy than elevation. But that's just me guessing. Am I anywhere near right?