How accurate are GPS's

Member - Tony V (NSW) posted:

Bravo,

Explorer is spot on there is an excellent explanation on datum and the differences at Australian hydrographic web site
It explains charts and the differents in WGS84, AGD66 or AGD84 and GDA94 etc.

Quote from page

"So how does this affect me?

If you have two charts of the same area and they have been drawn on different datums, then the latitude and longitude for the same place is likely to be different on each chart. In Australia the difference between an AGD66 and a WGS charted position can be up to 200 metres (0.1 miles). This could be the difference between one side of a reef and the other!"

Good info page with good links.

regards,

Andrew & Jen posted:

Hullo Peter
Setting aside the possible error due to an incorrect datum, GPSs experience "drift". Assuming you have a modern and sensitive unit, if you put the it in one place and observe it for about 20 or 30 mins, the coords will drift around by up to 40m.
One of the reasons for this is the fact that the GPS is continually calculating it's position by triangulation from a number of satellites, some of which are low down on the horizon while others are higher. Your position on the earth (not altitude) is more accurate if the majority of the signals are coming from well spread out, low down satellites.
If you have a tracking option and put it on the most detailed scale, after a while a pattern will develop like a multi pointed star and the centre of the star is the most accurate indicator of position.
If you need to accurately record a position, this is about the only way to do (unless you have a differential unit)
Cheers
Andrew

Steve63 posted:

The majority of the GPS network are rapidly moving satalites. There is also a set of satallites in a geostationary orbit (I think they have numbers over 40). The more satellites your unit can pick up the better. If it can also calculate the differentials from the geosyncronous satallites you will have less variance as the error effect is known. A lot of units will tell you if they have a differental on a satallite (WAAS). The more evenly spread the satallites are the better you are. Where the angles between satallites are very small the errors are more significant between those two satallites. Without differentials your error should be less than 10m and with differentials your error should be less than 3m.

Steve

Bravo Man posted:

Thanks for all the replies,I can see there are differences I didn't know about cause I'm an old fart and halve silly.

Ozhumvee....we came in from Mt Dare and did the loop east via the rig road and west by the French line back to Mt dare.Was very nice

Ozhumvee posted:

No worries thought that must have been the way you went, bet the desert was looking pretty green and lush I expect.
Re the GPS error when we first bought a Garmin GPS back in the late 80's it was a single channel unit but could multiplex to look at a max of eight satellites but of course one after the other. The error in those days was huge as the Yanks had the "scramble" on to stop the enemy from being able to use the system, errors of a few hundred yards were good nothing like now where a metre is commonplace. It all changed when Desert Storm started in the early 90's as the yanks didn't have enough Military GPS that could unscramble the signals so were forced to issue all the grunts with civilian GPS' which in turn meant they had to turn the scramble off and suddenly our GPS' would be very accurate. Initially they would turn it on and off when a mission was on and after a while it just stayed off as it is now.

muffin man posted:

Hey Bravo Man,
What were the big salt lakes like out poepel way. Had they dried and could you go through them or did you have to take the detours.
MM

Bravo Man posted:

Hi Muffin Man.
All the salt flats where dry but you could see tracks from earlier where they had been wet.

The only detour we had to do was the big one around Peopel Lake which had a fair bit of water.

The WAA track was blocked alltogether near the eastern end by a large lake.

Regards Peter