Technical explanation for why your GPSsucks at elevation (Read 850 times)

Very interesting.  But I don't buy that this explains most of the elevation variation I see in GPS.  Gravimetric bumpiness at the scale of an entire planet would not explain the GPS-derived elevation error that I see on a scale of a few hundred feet.  I would think it just has to do with problems of triangulation in a perpendicular plane reducing precision.

Yes, and it would effect barometric devices equally. 

Personally, I preferred my Loran backpack-watch.

Here are a couple of links to Differential GPS Aplications.

 

How does Differential GPS work?

 

Precision Agricultural GPS Solutions

 
Interesting.  I find it hard to believe that the agricultural auto steer, leveling, etc market will be anything but miniscule.  But it is kind of cool.

 

Do you have any idea what a tractor or combine cost these days?  They may not sell to the masses, but the price of an individual unit is hidden in the price of a multi-hundred thousand dollar unit.  Many farmers are much more high-tech than most people realize, and have been for years. 

 
Yes, I do.

It wasn't a comment on GPS technology versus the cost of equipment.  It was a market size comment.  And it wasn't a knock on farmers (most of one side of my family are or were farmers).  It's just that there aren't many farmers anymore.  And though the number of applicable farm equipment units has grown in recent years, the gross number is relatively small compared to, say, automobiles (or even runners).  But if it's a market that makes economic sense for the buyers and sellers, okay by me.

When we take a section of wilderness and turn it into a housing development or a road or whatever, we need to be very accurate both vertically and horizontally.  

The ONLY way we can be accurate enough is to use a differential GPS system.  We carefully calibrate a base station from usually 3-5 accurately surveyed points.  That base station generally knows where it is in the real world to within about half an inch.  (.02 feet is the best I've seen, .07 feet the worst)

That base station then listens to the satellites.  It compares the information it is getting and builds an offset to its real location.  It can be compensating for gravity, for atmospheric conditions, or for whatever.

The base station then broadcasts that offset to all of our machines. The machines apply the offset to the information they receive from the satellites, and are generally accurate to about an inch (vertically and horizontally.)  They lose as much as an inch of accuracy per mile from the base station.  

These offsets, especially vertically, can be hundreds of feet, even if the GPS receivers can "see" 7 or 8 satellites.

Some cities have networks of base stations setup on certain frequencies that can be used to get within a couple inches.  Reno Nevada has one, but it works on a different frequency than we use, so we set up our own base stations.

The construction setup is rather expensive... But, it has to handle the rigors of life in a bulldozer in the rain or the snow.  It has to be quite accurate (Generally, our contracts require about a 1 inch tolerance).  We are a smallish company but we have 4 base stations, 4 hand held rovers, and several machines.  

It seems that if some company wanted to, they could build car or wrist mounted GPSs that could see city wide networks if they were available.  Even being accurate to a foot or so would be better than being off 50 feet...  Its probably nearly as accurate and much cheaper to put in an altimeter of some kind though...

-Kelly