Major shift in DPOS processing

[Bruce Dawson]

Hello everyone,

I did a small job yesterday using my standard procedure when working in an area with no nearby control. I setup using an autonomous position with DPOS post processing. I always work in NAD 83/11 Washington North Zone. The position of my base was as follows (N,E,Z):

333337.72 1793105.65 1101.78

This was a reasonable location for the point, fitting with an ortho photo I had of the site. After DPOS processing I got the following results:

188216.801 1782563.120 2314.154

A MASSIVE shift! I did the processing twice and get the same results. To me this screams datum problem, but I have checked the project coordinate system on the LS and it is correct. I have done this many times and never had this happen.

Bruce

 

[Alexey Razumovsky]

Hello Bruce, share jps file with us.

 

[Bruce Dawson]

Okay Alexey, can I just do that through "Send Project to Support"? I know I have had issues posting to this forum with data before.

 

[Michael Glutting]

Hi Bruce, yes, that would be perfect if you submitted the entire project to support, Home/Collect/1.Project/Send Project to Support.

 

[Bruce Dawson]

On it's way.

 

[Vladimir Prasolov]

Hello everyone,

I did a small job yesterday using my standard procedure when working in an area with no nearby control. I setup using an autonomous position with DPOS post processing. I always work in NAD 83/11 Washington North Zone. The position of my base was as follows (N,E,Z):

333337.72 1793105.65 1101.78

This was a reasonable location for the point, fitting with an ortho photo I had of the site. After DPOS processing I got the following results:

188216.801 1782563.120 2314.154

A MASSIVE shift! I did the processing twice and get the same results. To me this screams datum problem, but I have checked the project coordinate system on the LS and it is correct. I have done this many times and never had this happen.

Bruce

Hi Bruce,

We sent partial data from raw data file to OPUS. The OPUS report shows:

After converting to US survey feet:

N = 57368.661 m * 3937 / 1200 = 188217.015 ft
E = 543326.441 m* 3937 / 1200 = 1782563.498 ft

That match DPOS report. There are small difference because we selected arbitrary interval.

I would suggest to verify base station position. It looks like it was input from different coordinate system.

 

[Shawn Billings]

Look at screen capture of base from base rover setup. See what the 2D and height delta shows in the image.

 

[Bruce Dawson]

I would like to Shawn, but where is the screenshot stored?

 

[Bruce Dawson]

Not sure if this helps, but here are some screenshots of info regarding the base point.

 

[Shawn Billings]

In the points screen you will see a screen capture icon. Highlight the base point and select the screen capture icon. When starting the base using base rover setup the LS creates a screen capture of the base rover setup screen.

 

[Shawn Billings]

You can also see the screen capture icon on the processed point information screen that you posted above.

 

[Bruce Dawson]

I did that Shawn, and the three I posted are the only screenshots. I remember hearing the "shutter release" noise when I started the base, but I can't for the life of me find it!

 

[Shawn Billings]

You might try Files, collected with point and collected by user.

 

[Bruce Dawson]

Yeah, no luck. I only have screenshots stored like the ones I show here. I went to a project done a few days prior and it had the screenshot of the setup, so not sure what happened on this job, but it seems to be gone. Or never existed.

Luckily I don't need the GPS data for this job. It was one that I ended up having to total station anyway (heavy tree cover), but I sure would like to know what went wrong. I know the project was setup in the right datum. The coordinates the Auto Locate gave me were correct for the area. I just don't understand where things went south. WAY south

Bruce

 

[Matt Johnson]

The coordinates the Auto Locate gave me were correct for the area.

I don't understand how you can justify the coordinates you had as being correct when DPOS and OPUS both give the same Washington North Zone coordinate that is totally different.

 

[Bruce Dawson]

Well Matt, I work pretty much every day in that datum. The general location I was working in is an area I have done several jobs, and they all resolved to a position similar to the Auto Locate position (Northing of 300,000). The DPOS resolved location is like 20 miles south of that area.

I also have ortho photos rectified to that datum, and the Auto Locate position fits for the ortho, the DPOS most certainly does not. The Lat-Long also fits for the Auto location.

 

[Bruce Dawson]

Okay, I may be on to something. Matt's comments made me look again at my stuff. The file I uploaded to DPOS had the correct date (3-13-18). It was named "1_090612" which did not correlate to what I thought it should be ("20_******" or something, I gave the base point a number of 20). I do not see any file with the date I was out there starting with 20. I also do not see the file when I bluetooth up to my Triumph-2 base. I know I downloaded it (at least I remember doing it) but it does not seem to be there.

So could the download have failed, and I sent the wrong file to DPOS, a file for a job 20 miles south? It applied a correction to that file, then shifted my points 20 miles south?

 

[Matt Johnson]

Yes it appears somehow the wrong raw data file got attached to your base point. 1_090612.jps has the following start and stop time:

Did you use "Attach GNSS file" with this point or do anything else out of the ordinary?

 

[Bruce Dawson]

No. And I remember specifically choosing to download the file and watching the counter click down.
Why don't I see the base file on the T-2? Does it delete it when you download?

 

[Matt Johnson]

Files remain on the receiver until they are deleted. Are there any files from the correct date on the T-2?