View Full Version : Discovery on Garmin's Altitude -

Wise Old Owl
07-20-2010, 09:26
I went to Delaware's Bethany shore to stay at my dad's shore home and was setting up a Garmin for a kayak trip. I noticed the house was at an altitude of -6 feet while standing outside... Thinking that was a bit strange I mentioned it to him and he said he had a flood plain survey done for insurance and he was 6 feet above flood plain. Google Earth data shows him at 6 feet above the back bay. So I recalibrate to the bay as sea level and drove back to the house only to get -6 again. Even if you leave it outside for 15 minutes - it won't change. Error on the unit I think was 15 feet horizontally. So I found this interesting. How does one really set the altitude with a good degree of accuracy?

The Old Fhart
07-20-2010, 11:24
I had posted this explanation of GPS altitude on WB a couple of months ago in another thread and have just cut-and-pasted it here. These comments pertain to using the GPS satellites to get altitude.

"the elevation a GPS displays isn't commonly understood. While a minimum of 3 satellites is needed for location, a minimum of 4 are needed for elevation. Signal strength and satellite geometry play a big part in elevation readings (http://docs.controlvision.com/pages/gps_altimetry.php). What makes matters worse is the elevation displayed on a GPS isn't the elevation above MSL (mean sea level) like a barometric based altimeter, but above a model of the earth which isn't exact. If you check out this NOAA reference (http://www.ngs.noaa.gov/PUBS_LIB/gislis96.html) it will state that if taken carefully, the GPS elevations could be about 175 feet off because the model of our earth isn't exact. The roatation of the earth causes it to 'bulge' in certain areas so it isn't an exact sphere. Elevation readings on maps have been taken by barometric based altimeters and when set recently at a known reference point, are generally quite accurate, generally within a very few feet."
You mentioned you reset the altitude on your Garmin so you probably have a 60CSx and were using the barometric based altimeter with the electronic pressure sensor. The specs say that should be +/- 10 feet when set to a known waypoint but I'd be surprised if it were that close. I'd say your GPS receiver is working the way it should and is within reasonable accuracy.

07-20-2010, 15:58
So your GPS said you were six feet under? Did you check your pulse? :D

Wise Old Owl
07-20-2010, 19:38
Thank's Ice age!

Wise Old Owl
07-20-2010, 22:23
Just matched the GPS to the Topo here at home in Paoli. The unit matches the altitude on the driveway. 360feet. Checked it twice over a half a day.

Wise Old Owl
07-26-2010, 18:36
Went back to the shore this week to see if I could do it again. Drove over to the kayak shop in the park and zero'ed it out again, went back to the house and it was +10 feet in the same spots - so I give up, even after leaving it for 10 minutes it never changed.

Great toy.

09-11-2010, 15:18
I'm usually pretty happy with 10' accuracy and don't expect to do better. Doesn't really matter if it is up, down, or side to side..... I still know where I am. 2dogs

09-13-2010, 13:48
My compass is pretty accurate. hasn't failed yet, and often comes to point accuracy of 0 feet..Thats the best I have seen.

I look up, look down, then I look all around... Then I look down again. yep, same spot. I am where I need to be at the moment.

This technique never fails me...And uses no batteries..

Old Hiker
09-13-2010, 14:04
You are lucky. My GPS keeps bouncing from 17 to 37 +/- feet, rapidly at times. I'm glad I don't get sea-sick easily!

Wise Old Owl
09-14-2010, 16:35
You are lucky. My GPS keeps bouncing from 17 to 37 +/- feet, rapidly at times. I'm glad I don't get sea-sick easily!

My understanding is the signal from the satellite is circular polarized from an early on reading when this was set up, the atmosphere causes the signals to flutter or fade in and out by moving the signal up and down the band slightly, I suspect the Land Surveyor stuff is left in one spot to make a cumulative calculation to get past the percentage of error.

Please feel free to add, because I might not recollect this correctly.