Corpscon 6 combined scale factor wrong?

>
> Can anyone describe under what circumstances Corpscon 6 returns erroneous values for the combined scale factor?
>
> Thanks.

you may want to check your corpscon results against ngs's nadcon. there is also good reads written about coordinate systems, one by joseph dracup 1977, the other by james stem in 1990

Have you checked their number against the real number?

I have not found the 'trigger' that causes an incorrect CSF. That being said I've never seen it appreciably incorrect. The error that shows up in our project areas is technically incorrect but practically sufficient. That is not necessarily true at a different elevation or location within a projection or zone. We work in search real States so rather than remember where it's good we just don't use it for generating a CSF. There are too many other good tools for the job.

I am using NAD 83 coordinates with NAVD 88 orthometric or NAD 83 ellipsoid heights taken from NGS datasheets as Corpscon input and comparing the Corpscon output to the datasheet combined factor. See e.g. AF7601.

The calculation is fairly simple, I'm travelling now so can't reference any publication, but you sould be able to tract it down then run the calculation for any point and see if corpscon matches. It's probably very minor. After the 6th decimal place it's mostly meaningless

This is an example from NGS datasheet AF7601. In every instance, Corpscon 6 is in substantial agreement with the datasheet.

[inlinecode]NAD 83 Latitude: 27 59 22.59392
NAD 83 Longitude: 081 39 39.32398
NAD 83 Ellipsoid Height: 17.391 m
NAVD 88 Ortho Height: 44.028 m
Zone: SPC FL W
Combined Scale Factor: 0.99995217

Corpscon 6 Results
Geoid CSF
Using Ellipsoid Height
All 0.999952173
Using Ortho Height
2012A 0.999952175
2009 0.999952175
2003 0.999952171
1999 0.999952169
1996 0.999952170[/inlinecode]

> This is an example from NGS datasheet AF7601. In every instance, Corpscon 6 is in substantial agreement with the datasheet.
>
> [inlinecode]NAD 83 Latitude: 27 59 22.59392
> NAD 83 Longitude: 081 39 39.32398
> NAD 83 Ellipsoid Height: 17.391 m
> NAVD 88 Ortho Height: 44.028 m
> Zone: SPC FL W
> Combined Scale Factor: 0.99995217
>
> Corpscon 6 Results
> Geoid CSF
> Using Ellipsoid Height
> All 0.999952173
> Using Ortho Height
> 2012A 0.999952175
> 2009 0.999952175
> 2003 0.999952171
> 1999 0.999952169
> 1996 0.999952170[/inlinecode]

Well, the Ellipsoid Height Scale Factor is obviously correct to eight places at 0.99999727. That's an easy one to check.

In the sense of distance between points 1 ppm is nothing. If your work flow involves use of CSF and 1/CSF it pays to go extra digits...

It's really pointless for any use I can think of. I've seen them carried out to 9 places but never for a good reason, however, once established then you're stuck with it.

If your data collector requires 1/CSF and office software requires CSF, 6 digits will generate coordinates with minute differences. Again no practical difference but it can cause problems with finicky software packages. If I can avoid the problem with 2 keystrokes per job I will...

Notwithstanding the Corps' warnings to the contrary, Corpscon 6 produces correct combined scale factors.

As Loyal Olson and others have reported, Corpscon 5 produces incorrect combined scale factors by using orthoheight (instead of ellipsoid height) to compute the ellipsoid reduction factor. To get the correct (Corpscon 6) result from Corpscon 5, use the ellipsoid height as the input orthoheight.

To get the Corpscon 5 (incorrect) result from Corpscon 6, use the orthoheight (with respect to Geoid96) as the input ellipsoid height. This presumably explains the Corps' instruction, "To obtain CSF, rerun the points with orthometric heights."

Corpscon 6 is an outstanding effort, even if the Corps is confused, and I am grateful to have it.

Discussion again in [msg=315602]later thread[/msg].

You said: Using Ellipsoid Height CF= 0.999952173

When I run Corpscon 6, selecting either GRS80 or NAVD88 meters and entering the ellipsoidal height 17.391 I get a combined factor of 0.999956352, not your value.

Entering 44.028 m (the ortho height) matches your numbers.

The datum for NAD83 ellipsoid heights is GRS80. (Using NGVD29 or NAVD88 makes the input an orthoheight.)

Maybe I nailed down the bug?

I had used GRS80 for both input and output. I just now tried the other combinations of input and output vertical datum.

Input            output         combined factor
GRS80  17.391    NAVD88 44.024  0.999952172  correct
NAVD88 44.024    NAVD88 44.024  0.999952172  correct
NAVD88 44.024    GRS80  17.391  0.999956351  wrong
GRS80  17.391    GRS80  17.391  0.999956352  wrong
Compare other conditions:
NAVD88 17.391    NAVD88 17.391  0.999956352  different input, gets the mystery value

It appears that the input may be selected to either datum, with the appropriate height value, but the output MUST be selected to be NAVD88 in order for the combined factor to be correct.

Is that how you see it? I would have thought how you choose to express the output elevation should have no effect on the computation combined factor, but it seems to.

It would appear that the bug is that it always uses the output elevation number to compute the combined factor, always treating it as NAVD88 regardless of what was selected.

Maybe I nailed down the bug?

Thank you, Bill. That answers my original question, "Can anyone describe under what circumstances Corpscon 6 returns erroneous values for the combined scale factor?"

I got a message today saying the Corps of Engineers newsletter for Surveying and Mapping Community of Practice linked this thread, thanked us for clarifying the issue, and recommended SurveyorConnect.

Corpscon 6.01 does exactly what I want.?ÿ Thx for the replies.