[APG Public List] [APG Members] place names
john at jytangledweb.org
Tue Oct 26 22:42:50 MDT 2010
What you say has truth in it, but in practice one might not
make the leap that the number of significant figures presented
means that the uncertainty is in the next decimal place.
In scientific publications one always attempts a computation
of RMSD (root mean square deviation) and write it with a +/-
to prevent confusion. Like 37.79507 +/-0.00005
Too many people just report numbers without computing their
uncertainty, so it is best to be explicit about it when you
mean it. You can never tell if the writer knows about
significant figures or not if only the number without an
explicit uncertainty is given.
In fact, I always record GPS coordinates to 5 decimal places.
I haven't done the computation to see if I should only be
reporting 4 decimal places. (decimal places are not the same
as significant figures). I rely on the uncertainty of
+/-15 feet that my device tells me, at usual best. If
fewer satellites are locked on by the device, it can go
to a much higher +/- uncertainty. My device gives me the
uncertainty, and reports to 5 decimal places, so I use it
as is. I always caution people that current GPS devices
(that I own at least) are only accurate to +/-15 feet at best.
Not good for locating the corners of a house. Probably, good
enough for locating the site of a house, existent or nonexistent.
Good enough for finding an existing grave stone in a cemetery
(quite like geocaching), but perhaps not good enough to place
a grave stone on an unmarked grave. Unless you document it
as the "locale" and not the "site" ;-) .
On 10/26/2010 2:29 AM, Stephen Danko wrote:
> GPS coordinates can be specified so that they represent either a large
> general area or a spot the size of a pinpoint. It's all in how many
> significant figures one lists in the coordinates.
> The number of significant figures refers to the number of numerals in
> the coordinate. 37.79507, -122.40280 has 7 significant figures in the
> N-S coordinate (the first number) and 8 significant figures in the E-W
> coordinate (the second number). This is the location of the Transamerica
> Pyramid in San Francisco. Actually, it is the location of a part of the
> Transamerica Pyramid. This coordinate has so many significant figures
> that it points to a very accurate location on the earth. If the
> coordinate is written as 37.8, -122.4, the area described becomes much
> larger and includes anything between 37.75 to 37.84, -122.35 to -122.44.
> Written as 37.8, -122.4, the coordinates don't accurately specify the
> Transamerica Pyramid, but instead specify about half of San Francisco.
> Note that 37.8, -122.4 is not the same as 37.8000, -122.4000. The first
> set of coordinates gives a general location (the eastern half of San
> Francisco) and the second set gives a very specific location (the corner
> of Vallejo and Front Streets in San Francisco).
> The latitude longitude system we use today is based on a prime meridian
> through Greenwich, England. Historically, different places used
> different prime meridians (including prime meridians in Paris,
> Washington, D.C., and Rio de Janeiro). For genealogical purposes, all
> historic systems can be ignored and we can use today's system with a
> prime meridian through Greenwich because, as Larry pointed out, our goal
> in using latitude longitude coordinates is to specify where a place is
> today, using today's system. Using GPS coordinates, Nicole (in a
> different thread) would have immediately been able to find out where
> Grossendorf, Germany was, and she would have immediately been able to
> see that the present name of the place is Władysławowo, Poland.
> True latitude/longitude coordinates are not exactly the same as GPS
> coordinates. This is because GPS uses the International Reference
> Meridian which is 102.5 meters east of the Prime Meridian through the
> Royal Observatory in Greenwich. Don't worry about the reason for this
> difference. Google Earth used GPS coordinates instead of true latitude
> longitude coordinates, presumably so that GPS systems will match up with
> Google Earth.
> In summary, GPS coordinates can be used to describe a very specific
> point on the earth or a very general area of the earth depending on the
> number of significant figures in the coordinates.
> Kind regards,
> Stephen J. Danko
> *From:* Connie Sheets <clsheets1 at yahoo.com>
> *To:* apgpubliclist at apgen.org; apgmembersonlylist at apgen.org
> *Sent:* Mon, October 25, 2010 4:33:55 PM
> *Subject:* Re: [APG Members] [APG Public List] place names
> I must admit to only a general knowledge about, and frequent lack of
> attention to, GPS coordinates. I have been operating under the
> assumption they describe a specific point on the earth, not a large area
> like a rural US township of 36 square miles.
> I can comprehend how I might want to visit the crossroads approximately
> one mile southeast of a very small village in Northwest Missouri where
> my great-great grandfather's house once stood, obtain the GPS
> coordinates, and record them for posterity with a photograph I have of
> the house. I also understand how GPS coordinates are useful for locating
> cemeteries, graves within cemeteries, and other landmarks.
> However, if all I know about an ancestor's location is a rural township
> or county, I would be concerned that I was promoting inaccuracy if I
> arbitrarily chose (and yes, it would be arbitrary) the center of the
> township or county.
> I will continue to use standard historical place names, with a reference
> to the modern place name when necessary, for the foreseeable future.
> Connie Sheets
More information about the APGPublicList