[San-Diego-pm] Meeting today (Monday, March 13th, 2006)

Tkil tkil at scrye.com
Wed Mar 15 14:50:20 PST 2006

```>>>>> "Joel" == Joel Fentin <joel at fentin.com> writes:

Joel> I'll switch my GPS to Lat-Lon and WGS84 and see if that cleans
Joel> off the error. If so, I will have gotten an eduction and I can
Joel> then continue with what I set out to do in the first place.

I ran a quick experiment last night.  Well, this morning.  Something.

1. Set GPS unit outside to acquire a fix.  It's in the default WGS84
mode, and came up with:

WGS84:         32°44.416' N  117°15.245' W

2. Switched GPS to NAD27 CONUS, and it came back with this fix:

Using the very rough approximation that a minute is a mile, that's
a difference of .05mi, or about 250ft, or some 80m.

Huh.  That's a better approximation than I thought!  I thought that
it was a mile-per-minute at the equator, but if we do:

length of equator * cos( 33° ) / ( 360 * 60 )
= ( semi-major axis * 2 * pi ) * cos( 33° ) / ( 360 * 60 )
= 6_378_137m * 2 * 3.14159 * 0.83867 / 21600
= 1_556.0m

So 0.05minute error is really close to 80m.

To get miles:

1556.0m / ( 1609.344 m/mi ) = 0.96686mi = 5105ft

Google Maps:   32°44.414' N  117°15.240' W

For comparison:

WGS84:         32°44.416' N  117°15.245' W
Google Maps:   32°44.414' N  117°15.240' W

So I'd say that google maps probably uses WGS84, or something close to
it.  By the same rough estimation above, the delta between WGS84 and
GM is an order of magnitude less than between WGS84 and NAD27-CONUS.
So GM vs. WGS84 at my house is only a 25ft / 8m difference, and that
was about the radius of uncertainty on my GPS at the time I took the
fix.

Mapping the NAD27-CONUS fix onto the GM datum puts me around the block:

http://tinyurl.com/prbyo

Joel> So much for Perl.

Well, I *did* have a link in there for a perl coordinate translator. :)

t.
```