We have GPS data from three separate days. Day 1 & Day 2 data differ by 0.25' horizontally and 0.1' Vertically. Both Day 1 and Day 2 was collected in "Beast Mode" until Javad accepted the position. With the 3rd Day we occupied the point for close to an hour and basically hit the middle of the two coordinates, only being 0.03' from the derived center of Day 1 & Day 2. The sky is relatively clear with some trees to the North East which I believe has influenced the difference in our horizontal coordinates. Has anyone experience similar results with coordinates differing that much? Our fear is that if we use Javad for Boundary work we may get a "bad" result without being confident in its position. Not all jobs allow for this much repetition.
Differences in Coordinates
- Thread starter ndoring
- Start date
More details are need about the settings used to collect these points. Can you attach or email me the project archive?
Was the base set up on the same point each day? If so, were the same coordinates used for the base each day? What profile was used to collect the points?
?
If you have several vectors that are supposed to be the same and the beginning endpoint isn't the same for all three vectors, then we can't expect the terminal endpoint to be the same.
I'm guessing to even get as close as you did, that each base was processed to CORS and the rover positions were then translated to the new base position. If the base sessions were fairly short then there will be higher error potential in the base position.
As Matt said, it will be better to see your project to be certain of exactly what happened. How do the inverses between the base and rover (using the base point that was used to create the rover point) compare?
Nick shared a link to the project archive with me: https://www.dropbox.com/s/j8pst82qq85g957/17185001GPS-171011.zip?dl=0
Which points are the ones you are referring to in your first post?
Which points are the ones you are referring to in your first post?
There seems to be some source of a lot of multipath at the location from looking at your screenshots. Notice the large spreads in the plots:
It looks like this is the position of it? https://www.google.com/maps/@38.010...4!1sp6QG4aBtIYkLGHNeHpYp1A!2e0!7i13312!8i6656
Is it close to the fire hydrant there? I see some kind of marker there that is painted orange but can't tell what it is.
Is it close to the fire hydrant there? I see some kind of marker there that is painted orange but can't tell what it is.
Nate The Surveyor
Well-Known Member
Inverse between all 3 base points. Draw a mini worksheet of this. Lable these 3 base points, day1 day2 day3
Inverse the 3 points in question.
Do the same as above. (Day 1, 2, 3)
I think you will find your answer.
N
Oops, others answered b4 me!
Youll get it.
Inverse the 3 points in question.
Do the same as above. (Day 1, 2, 3)
I think you will find your answer.
N
Oops, others answered b4 me!
Youll get it.
That is before they were corrected by DPOS. After DPOS all the base coordinates only disagree 0.013'. The vertical plots show drifting which is caused by multipath.
I can't say that I have ever seen that much difference when using my base. I do see it to some degree on the network I use. Usually if it's that bad of an area I usually don't get enough rtk to store the point and have to rely on static. I bumped up to 240 seconds. I think three minute mark is a risky place to end the observation.
Good question
John Thompson
Well-Known Member
If there were an indicator on the collect action screen somewhere that showed whether correct for tilts is on or not, it would be helpful to diagnose a situation like this. I think it would be helpful for use in the field too. I have been bitten a time or two when I thought it was on and it wasn't, or when when it was on and I thought it was off.
I agree
Tilts were off in his boundary profile when I checked it. Tilt correction wouldn't cause the behavior seen in the plots anyway. When I look at the location of the base it looks like it is in a less than ideal location, in a clearing in the trees behind this tall house:
Correction for tilts is off. I wasn't working in the field on this project but I agree that the base set up was not ideal. We often have to weigh security with our setup location and can understand why they chose to set it up there. When inversing the base station coordinates for each day I find that they are all plus/minus 0.01'. It interesting with the screen shots that the coordinates varied in opposite directions for Day 1 and Day 2, and Day 3 seems to vary down the middle. I suppose that just has to do with the satellite constellation.