[sac-forum] Re: Determining Position

Stan, ignoring both surface curvature and the effect of changing
latitude on longitude (both of which are minor over distances like
1000'), changes in lat/long have a linear relationship with changes in
distance on the ground. Therefore I don't see why you couldn't simply
set up a GOTO with the midpoint latitude value and midpoint longitude
value between the two known points. 

At our lat/long, 100' on the ground is roughly 1 second of lat or long.
So suppose your two known points were N 33 29 55, W 111 30 05 and N 33
30 02, W 111 29 57. The numerical midpoint is N 33 29 58.5, W 111 30 01.
I think this would be the midpoint on a straight line between the 2
known points, to the limit of GPS accuracy for all 3 points. (I've
approximated hypotenuse distance, of course, for this example. In
reality, you'd carry this another decimal place.)

This is in effect what your graph paper approach does. If you wanted to
go 1/4 of the distance between the two known points, you'd simply use
1/4 of the lat and long difference in setting up the GOTO, and etc for
other distances between the known points.

I'd say that once you get to distances of several miles or more, the
initial assumptions about ignoring curvature and latitude effects on
longitude cause errors that begin to exceed inherent GPS accuracy.

I've probably forgotten some important factor, but hopefully others will
catch my mistake(s).

Dan Gruber

-----Original Message-----
From: sac-forum-bounce@xxxxxxxxxxxxx
[mailto:sac-forum-bounce@xxxxxxxxxxxxx] On Behalf Of Stan Gorodenski
Sent: Thursday, May 18, 2006 9:02 PM
To: sac-forum@xxxxxxxxxxxxx
Subject: [sac-forum] Re: Determining Position

Since I am always thinking about something like this, it just occurred 
to me a way I could do it. I have plenty of old style graph paper, fine 
divisioned and some large sheets. I could set up x & y coordinates, one 
for latitude and the other for longitude, plot the two points, and then 
draw a straight line between the two and read off latituded and 
longitude for the points in between. Or, and probably the best, it 
should be easy to get a slope for a straight line just by the theory of 
a regression line. For my purpose, I do not need to take into account 
the earth curvature since any deviation would undoubtedly be much 
smaller than the accuracy of a GPS - After getting the in between points

I plan to go out to the field and locate them using a GPS.
Stan

Stan Gorodenski wrote:

> Assume I have the latitude and longitude of two points separated by a 
> thousand feet. I want to find the mid point on a straight line between

> the two, or for that matter, anywhere along the straight line. There 
> are trees between the two points so that they cannot be seen. What is 
> the best way to do this? Is there an easy way to do it with the GPS 
> units?
> Stan
>
>
>

Other related posts: