Good explanation, Matthew. I suppose that means that systems that can take advantage of the GROUND-BASED WAAS stations should perform a bit better on the elevation question because you have another point to calculate the triangulation/time measurements from. I have to believe that you would need a LOT of ground-based stations for that to work or the curvature of the earth would again get in the way of the ground-based signals (guessing here).
On Fri, Sep 13, 2013 at 9:32 AM, Matthew Hunt <[email protected]> wrote: > On Fri, Sep 13, 2013 at 10:13 AM, Darren Addy <[email protected]> wrote: > >> Elevation of location. (ASIDE: This seems to be where "all GPS units >> exhibit a weakness" but I'm not sure why that is. According to this >> PDF, >> http://www.nasa.gov/centers/johnson/pdf/584738main_Wings-ch5c-pgs360-369.pdf >> the space shuttle created topographical data sets that can give the >> elevation for virtually any location. Perhaps it is just not built >> into the GPS system yet, or there is no method for the GPS to "look it >> up" and enter it into the calculations.) > > GPS receivers, in general, calculate elevation the same way they > calculate position, from the difference in arrival times of the GPS > signals. While it is conceivable to use SRTM (the Shuttle data) or > another elevation dataset, I would be surprised of the O-GPS1 were > doing so. (If you hoist the O-GPS1 up a flagpole or something, you > should be able to tell.) Also, SRTM is itself somewhat coarse in terms > of the post spacing on the ground (30 meters or worse), so I'm not > sure if it would be an improvement or not. > > Why do GPS receivers do worse with elevation than with lat/lon? The > reason is geometric. When measuring position via arrival times, you > get the best results from having satellites all around you. If you're > measuring your latitude, you'd like a satellite to the north, and > another to the south. That ways if you move a meter north, you've > decreased the distance to the north satellite by a meter, and > increased the distance to the south satellite by a meter, and thus > changed the difference between them by 2 meters, which is reflected in > your difference of arrival time measurement. > > With elevation, the stupid earth is annoying in the way, preventing > you from seeing a satellite beneath you. They're all above you, and > thus somewhat correlated in the up-down direction. When you move up, > you move closer to all of them. They're not all directly above you, so > there is some variation to measure, but it's a harder measurement. > > There are various "dilution of precision" metrics that the GPS > receiver calculates to estimate these uncertainties based on the > geometry of the satellites at any particular time. I don't know that > the O-GPS1 makes those visible to the user, though. > > -- > PDML Pentax-Discuss Mail List > [email protected] > http://pdml.net/mailman/listinfo/pdml_pdml.net > to UNSUBSCRIBE from the PDML, please visit the link directly above and follow > the directions. -- Nothing is sure but death and Pentaxes. -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
