3 ms·
Given nearby local ground-based correction stations (e.g. WAAS) it can use and flat terrain, yeah, you can certainly get that kind of accuracy. In practice, th
by jofer 7y ago
Given nearby local ground-based correction stations (e.g. WAAS) it can use and flat terrain, yeah, you can certainly get that kind of accuracy.
In practice, though, I've spent a fair bit of time trying to re-occupy sites taken with handheld GPS receivers. (I'm a geologist -- it comes up a lot) 30m is a better representation of typical repeatability for a consumer-grade handheld unit or a phone.
You're often on the side of a hill, so you can't see the full constellation. If you don't have line-of-sight to more than a few satellites, you risk an inaccurate location. You're often in highly remote areas, so WAAS doesn't always help (it's based on proximity to a ground station). If you don't have WAAS, you're at the risk of atmospheric effects significantly affecting your accuracy.
There's also the user error side: You might have the handheld set to display in NAD27 one time and WGS84 another, which can easily offset the same lat/long by several hundred meters.
+/- 3m with consumer GPS is definitely possible, but that's for ideal conditions. You're rarely working with ideal conditions in the field.
- grawprog 7y agoThe best i've ever personally got with a handheld GPS under actual work conditions was +/- 15m. I've gotten +/- 5m in large open areas lacking trees, mountains or large rock formations but areas without any of those things are rare. Weather conditions can make a big difference. In stormy or rainy weather I was lucky if I could get +/-20m