On 4/3/2017 9:06 PM, Dennis Lee Bieber wrote:
> On Mon, 3 Apr 2017 17:48:21 -0400, rickman  declaimed the
> following:
>
>
>> I believe there have been a number of significant air accidents when the
>> navigation devices were not set up correctly or failed.  I recall a
>> sailing accident that cost several lives when the GPS constellation was
>> in a very poor configuration giving a very high error, on the order of a
>> tenth mile or so.  The ship hit the rocks and the obstinate captain who
>> wouldn't listen to the warnings of those who saw the rocks was killed
>> along with some others.
>
>  That degree of error probably also meant this was the days of
> "Selective Availability" -- when even a good constellation produced error
> circles of 100 feet or more (or is that reported in meters on my units --
> need to check the manual). But near major ports, even in the days of SA,
> there are differential GPS transmitters -- which do need extra receiver
> equipment...

You don't see to understand how GPS works.  It is triangulation based on
your distance from the various satellites.  If there are few of them
visible and if they are clustered close together and worse, if they are
all near the horizon, the math will not give you a highly accurate
solution no matter what augmentation you use.  This is called "dilution
of precision".  It has nothing to do with selective availability or
differential augmentation.  The only solution is to have more satellites
in the sky so the rare sparse constellations do not occur, which is what
they have done eventually.


>  Since SA was turned off, most GPS units are good for 30 feet (meters),
> and with WAAS/EGNOS, 10 feet (meters) or better is common.
>
>  WAAS/EGNOS transmit on the regular (civil) GPS frequencies, providing
> correction data for poor quality constellations. Differential GPS provides
> very detailed correction but only near the DGPS transmitter (basically,
> DGPS sends the difference between the GPS constellation results at that
> port and the surveyed location of the port). Most current production GPS
> decode WAAS (two satellites for west and east CONUS) or EGNOS (Europe).

The WAAS corrections have to do with fine corrections for satellite
orbits and clocks as well as atmospheric disturbance to the signals
which can not be easily corrected by other means.  They are also local
(relatively speaking) in that the calculations are performed for the
ionospheric distortions on the signal paths to your location.  None of
this will do anything to minimize the impact of a poor constellation.

Sat A - O--------------You
                         |
                         |
                         |
                         |
                         |
                         |
                 Sat B - O

Sat A - O--------------You
              __,,--''``
Sat B - O--``

The angle between the two paths determines the size and shape of the
error ellipse.  In the former case the error will be very close to a
circle with a minimum radius.  In the latter case the difference in time
of arrival will create an error ellipse that is very elongated with a
poor dilution of precision.  There is no method of correction.

https://en.wikipedia.org/wiki/Dilution_of_precision_(navigation)

--

Rick C

--- SoupGate-Win32 v1.05