[LINK] limits of technology in finding someone

[Gordon Keith]

On Sat, 9 Dec 2006 15:35, Stewart Fist wrote:
> What do people think about the accuracy of GPS -- in the best
> circumstances, and with the best gear.
>
> I remember when the USA had dithering on the system (to deliberately reduce
> the accuracy to a few hundred metres, maybe even more).  At that time they
> boasted that their military accuracy, with no dithering, was a matter of
> centimetres.

My understanding is that GPS transmits on two bands, one public and one 
encrypted. The public band also used to dithered, but isn't any more. To read 
the encrypted military band you need special, classified, equipment.

The accuracy of both the public and military bands is significantly better 
than just that of the public band. Although I understand it's not as good as 
good differential GPS.

Differential GPS comes in a number of forms. 

Short range DGPS, where you have a fixed base station near the area you are 
working and continuously transmits differences between current GPS position 
and "known" position (usually long term averaged GPS position) is probably 
good for centimeter accuracy, depending on conditions (range from base 
station, etc). This can eliminate errors caused by local atmospheric 
conditions as well satelite errors

Long rang DGPS services, where a number of base stations analyse signals 
received from the satelites and transmit correction information are good to 
within meters. These are used by the aviation and maritime industries (my 
background is marine research). These correct for satelite positioning 
errors, but not for (upper) atmospheric effects on radio propogation. 
Accuracy is also better near base stations, and not as good well out to sea.

Uncorrected undithered GPS is good for tens of meters. Jumps in position of up 
to about 40m are not uncommon and stationary objects will "move" at up to 
about 2 m/s.

Uncorrected dithered GPS, like the bad old days, is good to within a couple of 
hundred meters. Data from a fixed GPS unit was quite adequate for testing 
plotting software.

> In fact, I distinctly remember one claim that it was measurable in
> millimeters, and that it was more accurate for house-block surveying than
> the conventional terrestrial triangulation methods.

With a good short range DGPS system I would believe it.

> The question of exaggerated claims to accuracy in these matters is
> interesting because of the claims to millimeter accuracy in measuring the
> sea-levels from satellites.  The satellites are perturbed by gravitational
> anomolies they pass over, and they are pushed around by solar winds, and
> they are flying over an Earth which is spheroidal measuring the distance to
> water which has a surface changed by waves, surges, tides, atmospheric
> pressure lifts, etc. and land masses which are rising and falling tidally,
> as well as tectonically.

You forgot about the affect of the atmosphere on the path of the radio signals 
which is also significant at that level of accuracy.

But many of these errors can be eliminated by back calculating the satelite 
position from known base stations. 

> I just can't see that they could possibly have accuracies in their
> satellite sea-level measurements better than a hand-span or so -- but I've
> never been game to say so in print because I don't have any direct
> experience or evidence.

For things like satelite sea-level measurements GPS is used to measure the 
position of the satelite by triangulation against a number of calibrated base 
stations and radar is used from the satelite to measure sea position.

The system also doesn't just throw out a number in real time, like a hand held 
GPS. The positioning data and radar data are processed to get results that 
are statistically robust.

For some old RADARSAT images:
http://www.marine.csiro.au/~cresswel/RADARSAT/ from 
http://www.marine.csiro.au/remotesensing/

Regards
Gordon