GPS accuracy

[Guest TSewell]

I'm brand new to this. Are the less expensive, $!00 - $200 units going to get within a few feet like I saw on TV?

Recommended unit(s) for a beginner?

[Guest OregonSurveyor]

In my limited understanding, I believe essentially all models use the same software to determine position. Some may have higher quality antennaes and/or the ability to connect an auxiliary antennae. The differences in price are more related to waterproofness/durablity, downloadable maps, and "bells & whistles". Under the most favorable conditions (9 to 12 satellites in a nice distribution) you are kidding yourself if you think you are getting better than 10 foot accuracy. That's not to say you can't been within 3 or 4 feet of the true coordinates, but you can't do this repeatedly and with predictability. Generally accuracies are in the 20 to 40 foot range, but this is a loaded question.

I like my good old Garmin GPS12 ($140 range), but you can't down load maps. In short... price does not buy you accuracy.

 

------------------

Lat. N 45d-30.6875'

Long. W 122d-37.2055'

Elev.=172.97'

king6kids@hotmail.com

[Guest swangner]

Theoretically, the best accuracy you can get with a GPS only unit is 3 meters (approx. 10 feet). This is under the best of circumstances - under normal circumstances, expect more like 20-30 feet. The only way to get better accuracy than this is to get a differential GPS (DGPS) unit, which combines the GPS signals from the satellites with signals from ground stations. The downside is that these units are VERY expensive. DGPS units are the ones used by surveyors and people like that.

 

As far as price vs. accuracy, higher price does not give you more accuracy. It might give you more features, but any well-designed 12 channel parallel receiver GPS unit (which most are) will give you roughly the same accuracy.

 

[This message has been edited by swangner (edited 02-12-2001).]

[Guest trilobites]

The question of accuracy has come up in several forms at several times. Thus, over the last month I conducted a test of an Etrex Summit at a USGS bench in front of our USGS office in Rolla, MO [MCMC76].

Readings were taken at various times of the day between 0600 and 1830 with a sample size of 80. The Etrex was warmed up for 1.5 minutes and held at random compass headings (if you hold it the same way each time a notable shift if the readings is observed).

 

Actual coordinate is 610270, 4201245

avg Etrex reading was 610269.4, 4201246.5

with std deviations of -0.6, and 1.5

 

What does this mean? From Quantile plots...

5% of the readings were within 1 meters

25% of the readings were within 3 meters

50% of the readings were within 5 meters

70% of the readings were within 6 meters

80% of the readings were within 8 meters

95% of the readings were within 13 meters

 

Thus, the positional error is just about what the manufacturer states. The site has several trees to the west and a brick building to the east. Most of the time between 4 and 7 channels were locked and the display generally indicated an error of 14 to 22 feet. Based on this "quick" test, the error listed on the Etrex screen is a pretty good approximation of the actual error, but forget about getting within a few feet most of the time!