Colorado accurate to 1/10th second

[gallet]

I just ran into this yesterday so I had to come back today and check my gps against it.

 

 

 

[Bill93]

You might want to look into what the difference is between WGS84, which is the GPS native datum, and the Australian datum, and how accurately the Colorado model reflects that difference. In the case of the North American Datum of 1983 versus WGS84, most recreational grade GPS units ignore the approximately one meter difference.

 

There is nothing else in a GPS to get out of calibration. If it works, then it works as designed and will not drift off from what the received signals tell it.

 

The main issue with accuracy is that signal conditions will give you a wandering position value. The real test is how often the unit is within some tolerance of the right reading. I see that your unit is expressing a 3 meter confidence (about 0.1 second latitude), probably meaning that it thinks there is a 95% chance the reading is that close. It is saying there is one chance in 20 that it is off further than that. That is typical for open sky conditions.

 

That will be very close to the statistical accuracy considering only the signal versus radio noise and local signal reflections. It ignores general slow fluctuations in the ionospheric path from the satellite that often are another 10 meters and sometimes more even if you have a differential correction like WAAS.

 

If you want to study your unit's accuracy in more detail, I would recommend working in decimal degrees instead of deg-min-sec, because most units give you better display resolution (smaller distance for least digit change) in that format.

 

You may also be able to find data on other geodetic marks from the Australian equivalent of the US National Geodetic Survey or a some regional government. Usually this data is very accurate for either the elevation or for the horizontal position in older marks, or both in newer ones measured with professional GPS. Most recently taken data is within a few cm, or 0.001 second for horizontal position.

Edited September 5, 2011 by Bill93

[AZcachemeister]

Regardless of the other details, that is an AWESOME monument!

[+BasicPoke]

FWIW, from my experience making electronic maps, the difference on the earth between a WGS84 position and a NAD83 position is much less than 1 m. I believe it is may be under 1 mm. We always treated them as equal.

BasicPoke

[Bill93]

You can use this tool to calculate the difference between NAD83 and WGS84 (choose the most recent flavors of each) for any given point in the US at any given recent date. One meter is typical, and the velocity between them is on the order of 1 or 2 cm per year.

http://www.ngs.noaa.gov/TOOLS/Htdp/Htdp.shtml

[DaveD]

Bill93 is correct. While the relationship varies a bit across the country, the difference between NAD 83 and WGS 84 is approximately 1 m each in the horizontal (latitude and longitude) and ellipsoid height components. People often mistake the difference in the reference ellipsoids, Geodetic Reference System 1980 (GRS80) for NAD 83 and WGS 84 for WGS 84 (a little confusing I know) which are sub mm across the North American continent, as the difference in a position realized in the respective reference frames.

[gallet]

I went back to check the difference between GDA94 and WGS 84 According to the SSSi website the Australian plate moves about 7cm per year.

 

GDA S27.22590 E153.11526

WGS S27.22590 E153.11528

 

However I only averaged the waypoints for a minute, which was two samples.

 

I usually do have my gps set to decimal degrees and only set it to d,m,s because that is all that was on the benchmark, which is odd when you think about it especially as Australia is a decimal country and seconds are pretty well redundant.

 

In the 'location (lat lon)' field which is one of the data fields on the Colorado, it always display as dddºmm.mmm while the 'location (selected)' displays what is set in preferences>position format. So it's not possible to directly compare dddºmm'ss.s" with ddd.ddddd which is kind of annoying because the default dddºmm.mmm is neither fish nor fowl.