I set two controls points, tied to section & quarter corners, and set a check point along the perimeter of a mile and a half boundary survey. Check point had some cover, solutions processor said the two vectors to it failed with a partial solution. So, I said what the heck, it's a check point, so I adjusted the data.

My control points on the highway check within a couple hundreths, so I take off through the woods. I tie to my check point with over 9' horizontal and 1' vertical bust. I didn't really care about the vertical, but I thought that I'd forgot to zero a backsite. So, this morning, I set up on the point that the perpendicular bisector pointed closest to, and the angle checked. Move on down to where I quit yesterday scratching my head and do a sun shot. Azimuth checks within 1 second, which has to be a fluke using a 6 second gun and the calculator almanac. I continue staking my lines through the woods, close about 1:10,000.

The moral of this story: If ashtech solutions says that the vectors failed, they may not be within three meters, partial solution or not.

terry

john francis and i are

seeing similar results.....

this is the fun part of opening the box and jumping out of it! it is cool to be able to conventional between yer points...to give you that warm fuzzy feelin'. to have the time to do so also ::)


i love my weekends now.... i get to speriment with the blue dillyjobies :)



ty

I know that everybody gets bored with me asking folks to Read That Fine Manual but the section in the Ashtech Solutions manual titled Post-Processing Data Analysis has a good discussion about partial and float solutions.

My personal preference is for fixed solutions. Sometimes those are just plain hard to come by. Sometimes a partial or a float solution is as good as it gets.

A skeptical attitude when it comes to accuracy is a good thing. Many of us remember wearing out a chain trying to decide whether the EDM was worth the money.

Please hear me strongly suggest that partial and float solutions should be checked by redundant measurements. But most of us already know the three words that guarantee accuracy with any survey measurement.

Does this mean you have to stay inside the box and pivot steer at the corners? No way! It's your dime. The only time I have a problem with outside the box science projects is when somebody wants me to explain why it did not work.

But a partial solution 'can' be like a partial car payment..and we know how much the bank likes those.

I've used Floats on very long lines at times, but I new what was happening and why it was happening (L1 stuff)..

I suppose partials can be okay, but I'm not comfortable with them for survey work. They indicate cycle slips and other data damaging problems.

Deral

Deral, I agree about 'floats'; The surveyor must have a good handle about what is going on with these observations in a network before accepting floats as part of the adjustment. I separate them into a different observation group for error estimation and weight.

Whether or not to remove them from the adjustment depends upon the processing results AND the adjustment stats AND common sense that comes from experience AND the project accuracy specifications.

I'll use them for well surveys out in the remote big timber if they meet tolerances, but not likely for other survey work.


Modified By Scott Partridge on 9/2/2004 at 7:02 PM

I've used partial and float solutions before on occasions also. If it had been something besides a check point, i would have re-observed or trashed it. I was just suprised at such a large error with short vectors (less than a mile). The way I was reading the data, i figured a foot or so at the most. I've been using the locus/promark/soultions for three + years, and this venture has made me maybe a little wiser and definitely more prudent.
I wanted to re-load solutions 2.6 and process and see if there was any difference, but I just don't have the time. I'm spending all my spare time occupying points fer that little extra edge.
Patience is a virtue, and time is not on my side.

Modified By J. T. Strickland on 9/2/2004 at 7:35 PM

Sometimes floaters can mean the things that you do not want to see in a swimming pool (if you catch my drift)..
Other times they are fine.

Depends on the length of lines and other factors. Mostly experience will tell you what is good and what is not when processing.I also use component groups like Scott...

A good indicator and also a good way to not bias the actual fixed positions by using the floats as the same weight.

Maybe I don't actually know what a partial solution is? Please define this.

My experience is not with Solutions but with another brand. Thanks

Deral

i guess you got to know the difference.


in gps talk and processin' i dont. but!

when it comes to swimmin' i have swam away from a few floatin baby ruths, in my time.

good discussion guys...

ty

Hey, is a 'partial' solution like being 'partially' pregnant? Or a 'partial' Republican or Democrat. Or like being 'partially' Catholic like my wife says she is...LOL.

Seems to me and I am sure TM knows more about this than I do when you get into long baselines L1/L2 50 Miles plus, all you ever get is Float Solutions, but with long occupation times they are supposed to be correct.(I think)

Take a careful look at the common satellites used in that partial solution. Then check you satellite data noise, etcetera. You can sometimes improve a solution by removing a crappy satellite. You can sometimes improve your position calculations by removing a vector from two nearby observations that do not have enough common satellites. It can go both ways.

I worked way outside the box in my wall experiment in Queens this past winter. Search for the whole report. I got a much better solution by locating the position of two observed points independantly from CORS sites. the points where 200'± apart and had a clear line of site. Independant position solutions were within 0.2' of the traverse measurements. Allowing solutions to solve them both at the same time gave a 20'+ difference.

For this one I got to tell Phil what I learned, and explained to him how I got through the quagmire. Still wouldn't have finished it without Phil. I have always got my money's worth on every 800 call.

Paul in PA