Community Forums

Carlson SurvNET GPS Adjustment  

  RSS

NeilRick
(@neilrick)
5+ posts Member
Joined: 1 month ago
Posts: 11
January 17, 2019 2:12 pm  

We recently did a survey through a Base/Rover procedure using a localization on two state place coordinates.  Then went through the adjustment process in SurvNET through Carlson Survey and adjusted the closure.  After realizing we had left out some road shots, we went back and shot in the road using the State Plane Mifi system without the Base set up.  When I dropped in the coordinates in Carlson there is an offset jag where the two road shots should meet, about 6 foot.  Is this because we used the Base on one set up and the Mifi on another?  Shouldn't it all be tied into the State Place coordinates anyways, or was there an error in our process?  Any one have any experience with this?


ReplyQuote
MightyMoe
(@mightymoe)
5,000+ posts Member
Joined: 8 years ago
Posts: 5830
January 17, 2019 3:36 pm  

Two state plane coordinates? From where did you obtain them?


ReplyQuote
Dave Karoly
(@dave-karoly)
10,000+ posts Member
Joined: 9 years ago
Posts: 10021
January 17, 2019 3:47 pm  

Did you press "here" for your Base position on the first survey?

O death, where is thy sting? O grave, where is thy victory? -1 Corinthians 15:55


ReplyQuote
NeilRick
(@neilrick)
5+ posts Member
Joined: 1 month ago
Posts: 11
January 17, 2019 4:32 pm  

We located with state plane using a wifi connection and mifi set up in our state.  We were tied into the local state plane and shot the GPS coordinates with the mobile Rover.  We shot each point twice and averaged the points.


ReplyQuote
NeilRick
(@neilrick)
5+ posts Member
Joined: 1 month ago
Posts: 11
January 17, 2019 4:34 pm  

When we set up the base it was not over a known point.  We shot in the base using satellites.  We then localized on the two mifi coordinates we had set.


ReplyQuote
Bill93
(@bill93)
5,000+ posts Member
Joined: 9 years ago
Posts: 5160
January 17, 2019 4:52 pm  

I guess I posted on the wrong duplicate thread.

Did you use a Combined Scale Factor to scale your total station distance measurements to state plane?  Failing to do that will give serious residuals in most areas.


ReplyQuote

Javad GNSS

NeilRick
(@neilrick)
5+ posts Member
Joined: 1 month ago
Posts: 11
January 17, 2019 4:58 pm  

We didn't use the total station in this survey, only GPS.  I know, confusing, two posts.  Two different jobs.  The other post is about the combination of the total station and GPS files.  I replied to that post.


ReplyQuote
Norman Oklahoma
(@norman-oklahoma)
2,500+ posts Member
Joined: 4 years ago
Posts: 4155
January 17, 2019 7:53 pm  

You are being asked about the source of your base position because 6 feet is a typical distance between an autonomous position and a corrected one. And even if your base position is correct, rover positions that are not "fixed", and are therefore autonomous, may be recorded.  Another possible source of error is a mixup with international foot / US Survey Foot settings.  

This post was modified 1 month ago by Norman Oklahoma

"Convention is like the shell to the chick, a protection till he is strong enough to break it through." Learned Hand


ReplyQuote
Nate The Surveyor
(@nate-the-surveyor)
5,000+ posts Member
Joined: 9 years ago
Posts: 8499
January 17, 2019 8:16 pm  

DGPS that's your answer.

Differential GPS.

If you set up a lone gps, 100x, and store data, on the same point, for 2 hrs each time, (shut it down, and restart each time) you will wind up with 100 coords, inside a circle, of varying sizes, but typically having a diameter of 10-15 ft.

What do we learn from this?

That gps, by itself is good to 10 or so feet.

Now, set out 2 receivers, and process together, and we have DGPS. It's good at determining the DIFFERENCE between two receivers.

Now, repeat this setting out of 2 receivers.  100x. The difference between them, (2 hr observations) is quite good. But, now both receivers are bouncing around 10-15 ft, but they bounce together. 

What do we learn from this?

We need other receivers, such as opus, cors, or a known point, with a receiver on it, to "stop the bounce".

DGPS. That's your answer.

The satellites only give you 10-15 ft.

There's your 6 ft.

Nate

Surveying is more than a Job----it is a passion to provide a foundation for future generation, that is beyond reproach.

Nate


ReplyQuote
Jitterboogie
(@jitterboogie)
250+ posts Member
Joined: 11 months ago
Posts: 259
January 17, 2019 8:29 pm  

I'm confused. If you used a mifi(VRS) RTK solution, As Norman mentioned, the shots should be more accurate than 6 ft unless you were in some really crap canopy or deep canyons with poor horizons. Base not being on shouldn't be that much of a factor because you have a solution for the new data on the new satellite constellations for that epoch. I say go shoot them again with the VRS, make sure you're showing fixed positions with good residuals, and compare the new shots.


ReplyQuote
NeilRick
(@neilrick)
5+ posts Member
Joined: 1 month ago
Posts: 11
February 1, 2019 3:38 pm  

Sorry.  To get back to this 2 weeks later.  I think the problem with the data is with the adjustment process in SurvNET.  When I simply put in the unadjusted coordinates into the drawing, it all matches.  In that case, the Base/Rover and MiFi shots will line up.  But if I adjust my Base/Rover shots using SurvNET and then try to drop in the MiFi shots, it jags.  I guess I am unsure what the adjustment is doing to the vectors to cause this to happen.


ReplyQuote
James W Johnston
(@james-w-johnston)
5+ posts Member
Joined: 2 years ago
Posts: 16
February 10, 2019 12:26 pm  

It relates to the old "precise vs accurate" consideration.

If the base in the base rover combination was properly localized to the mi fi points  than your base/rover measurements should be reasonably precise because of the short baseline, but not highly accurate because they are based on two points with a longer baseline.  It's not reasonable to see a 6 foot residual.

But if you used "here" or an averaging method to determine the base position the base/rover points will only be as accurate as if you have set up a single GPS receiver...about 10 to 15 feet.  But precise within the overall network.  A residual of 6 feet would be reasonable if comparing to the mi-fi points.  It would be a consistent residual over the whole network.

When you described your procedure...shooting two points with mi-fi, then localizing the base/rover points to those two points and continuing the survey, it sounded like you were doing the right thing.

 

James Johnston
B. Ed, Dipl. Geomatics


ReplyQuote

Take the next step in your career, pass your licensure exam with PPI2Pass

Ladd Nelson
(@ladd-nelson)
500+ posts Member
Joined: 9 years ago
Posts: 614
February 12, 2019 4:51 pm  

Trying to follow what I think was the chain of events...

  1. Posted by: NeilRick
    We located with state plane using a wifi connection and mifi set up in our state.  We were tied into the local state plane and shot the GPS coordinates with the mobile Rover.  We shot each point twice and averaged the points.

  2. Posted by: NeilRick
    When we set up the base it was not over a known point.  We shot in the base using satellites.  We then localized on the two mifi coordinates we had set.

  3. Posted by: NeilRick
    We didn't use the total station in this survey, only GPS.

  4. Posted by: NeilRick
    After realizing we had left out some road shots, we went back and shot in the road using the State Plane Mifi system without the Base set up.

I believe the secret here is item #2 "When we set up the base it was not over a known point.  We shot in the base using satellites." Since the base was likely over an arbitrary unknown position and in an Autonomous GPS solution (typically with a 2-meter accuracy range), I suspect it is accounting for the discrepancy you're describing because SurvNET is examining the processed results of raw file observations (e.g. lat/long/ellip. elev.) against your specified coordinate system rather than  "localized" results. Working backwards:

  1. The road points shot with step #4 should match the local control shot with step #1 above because you used your wifi connection and mifi set up in your State in both situations.
  2. The base position shot with step #2 was (say) Autonomous so points collected from it wouldn't normally match anything other than itself.
  3. To correct this, you localized the first round of data for the job on the local control shot with step #1 above.
  4. The result is:
    • Processed coordinates that should properly align, but,
    • Positions that have been derived from two different GPS base station locations (a network provided base and your autonomous base)
  5. Presuming that you let your field software derive the State Plane coordinates shot in step #1 (in other words, non-localized), the mixing of localized and non-localized data is probably at the root of the problem.
  6. I'll defer to those here with the much more vast field and practical experience than I but one recommendation would be to consider (if practical) the placement of your base on one of the two State Plane coordinates shot in step #1 after they've been located or locating a temporary State Plane coordinate for the base using the same methodology as you used in step #1. This would help minimize the error you're seeing.

However, all is not lost (at least I don't think it would be). If you have Carlson Survey, you might explore the Survey -- Edit-Process Raw Data command. Import the raw file. Once imported, explore the Process (Compute Pts) -- GPS routine. With this routine, you could explore the option of re-incorporating the localization option you might have used in the field.

I hope this information helps.

--
Ladd Nelson
Carlson Software


ReplyQuote
Take the next step in your career, pass your licensure exam with PPI2Pass
  
Working

Please Login or Register