In reply to ianstevens:
> Either the OP knows some deep, dark GPS knowledge that has evaded academia and the rest of society, or its a load of codswallop. As "I'm only 15 years old", I'm inlcined to think the latter. I don't want to seem like I'm picking on the OP who is clearly intelligent (I couldn't get an app to work) rather that his/her understanding of GPS and accuracy/precision is flawed.
> GPS hasn't had built in error since 2000, when the error signal was switched off and GPS became mildly useful to the general public. Unsuprisingly, this is when SatNavs and handheld units entered the market. Improvement in hardware at the non-satellite end of the system means that precision can be as low as 3m in x and y for a device such as your phone with a dedicated GPS chip. For most of us, that's plenty. For lower precision (i.e. "less error", two units need to be used in tandem in correlation with a "known point".
> I beleive the military now use a different system. In addition, Russian (GLONAS) and European (Galilaeo) systems are also being developed/launched as we speak.
It might be possible to integrate the accelerometers, and combine that data with GPS locations into a probabilistic model (Kalman filter maybe) to eke out a bit more accuracy - but that would be rather difficult to do (it would easily be a university-level project). Secondly I vaguely recall hearing that mountains/valleys can produce echos that will result in much bigger accuracy problems - a quick look shows these to be called 'multipath effects'.