Re: Converting co-ordinates from one datum to an other?

jimirwin wrote:

Thanks, I think Iv'e found a good one: http://www.colorado.edu/geography/gcraft/notes/datum/gif/molodens.gif

I also found an online calculator: http://www.ngs.noaa.gov/cgi-bin/nadcon.prl

I tried Peter Dana's version of Molodensky's formula and his example and
it works. I then tried it on NADCON and got slightly different answers,
30.0002238973/-100.0003696013/0m, 30.0002238891/-100.0003695878/232m (Dana)
and 30.0002112611/-100.0003501667 (NADCON). So changing the height
doesn't change it much.
So I then tried different longitudes based on a latitude of 40:

latitude longitude
NAD27: 40.0000000000 -65.0000000000
Dana: 39.9999643708 -64.9992930771
NADCON: 40.0001671027 -64.9992685694

40.0000000000 -70.0000000000
39.9999915613 -69.9994472150
40.0001355250 -69.9994378722

40.0000000000 -80.0000000000
40.0000255521 -79.9997669068
40.0000737694 -79.9997744055

40.0000000000 -90.0000000000
40.0000315820 -90.0000936811
40.0000484805 -90.0001139222

40.0000000000 -100.0000000000
40.0000094676 -100.0004176089
40.0000058944 -100.0004058361

40.0000000000 -110.0000000000
39.9999598811 -110.0007288478
39.9999636277 -110.0007048388

40.0000000000 -120.0000000000
39.9998843289 -120.0010179411
39.9999048055 -120.0010352250

40.0000000000 -130.0000000000
39.9997851067 -130.0012761048
39.9998735527 -130.0013791138

I understand these use different methods and NADCON uses actual recorded
data values in its calculation, so it is probably more accurate.
But Dsna's formula is probably simpler.
The next question is is there a formula for finding Deltas X, Y and Z,
perhaps a way to make them variables, thereby increasing the accuracy at
the global level?
If not, is there a table somewheres that gives the different values
between the different datums?

...

There is a subtle point here. NADCON is the US legal conversion
standard --even though it is based on certainly some mismeasured points. So
you need to define what is meant when you say accurate. If you think about
it, they needed to do that as one does not want to be arguing old property
line.

But generally the only recommeded accurate way to locate a point in one
system is to measure it in that system. If converted to another system, it
is best to state which transformation was used.

You should also get the following 2 excellent pieces of software
CORPSCON (a windows implementation of NADCON)
http://crunch.tec.army.mil/software/corpscon/corpscon.html
and GEOTRANS.
http://earth-info.nga.mil/GandG/geotrans/index.html


There certainly are talbles of the deltas. Run, do not walk to the following
websites
http://earth-info.nga.mil/GandG/coordsys/csat_pubs.html
http://earth-info.nga.mil/GandG/coordsys/datums/index.html


But keep in mind that these should be take as approximations. In addition
there is some ambiquity in some of the definitions. Through on top of that
that land masses drift--and by different amounts so things are time
dependent transformations in some cases.
Make sure you read the red at
http://earth-info.nga.mil/GandG/coordsys/datums/notice_8142002.html

I meant to also include in my post:

If you want to experiment with the difference between WGS84 and NAD83,
download

and also get HTDP from
http://www.ngs.noaa.gov/TOOLS/Htdp/Htdp.html

But also be certain to see Section 7.1 of
http://earth-info.nga.mil/GandG/publications/tr8350.2/wgs84fin.pdf