I'm trying to figure out both the angular and linear error of closure for a closed link traverse. We started out using 2 RTK-observed points. The first one, point 150, being a 1/4 corner. We closed on 2 RTK-observed points. I get a bearing of S 05?ø50'19" W between the closing traversed points (167, 168), and a bearing of S 05?ø49' 59" W between the closing RTK-observed points (169,170). My angular error of closure is therefore 0?ø00'20". Would my linear error of closure be 0.559' (167-169), or 0.495' (168-170)? Would the distance traversed be from point 151 (my first occupied point) to point 167 (my last occupied point)?
Is there a minimum technical standard for a mixed traverse using a total station and RTK? I didn't see anything in the Florida MTS. Please see the attached text file for the traverse data. Also, please see the attached CCR for point 150. Should we have used the CCR scale factor as the job scale factor?
I just realized that this traverse file isn't right. The points aren't in the proper order because I don't have the field notes. I will repost tomorrow evening. Sorry.
?ÿ
?ÿ
Maybe I'm missing something, but there is a list of points and a Corner Record, I don't see any traverse data.
No you don't need to use the same scale, there isn't anything there to use it for.
The lat, long and coordinate data will tell you if the coordinates are SPC, and it's NAD83(90) which may or may not be very close to a present day coordinate using OPUS or CORS.?ÿ
Maybe I'm missing something, but there is a list of points and a Corner Record, I don't see any traverse data.
I can't get the raw data file, but I can get the field notes. I will post the angles and distances tomorrow evening. I will also repost the point file containing only the points relevant to the traverse.
If you are checking a total station traverse against RTK measurements, you need a Combined Factor applied one way or the other so that both are at grid or both at ground. Rather than use someone else's factor at one single point, I would average the CF values for my RTK points.
You also need the convergence angle at each end taken into account.
As mentioned, NAD83(90) coordinates may be measurably different from NAD83(latest).
There are multiple ways to sort this out. You could fix points, fix azimuths, some combination thereof, run a least squares analysis. You could apply a CSF to the entire project, or CSFs to individual stations/observations. If you are trying to tie to published values and only have an old realization/epoch, you could transform the coordinates of that station(s) and adjust, or first transform your GNSS observations. It's all project-dependent and/or supervisor-dependent.
You also need the convergence angle at each end taken into account.
This is absolutely true if you are fixing to geodetic azimuths at either end. If you are using grid azimuths, you won't need to worry about the convergence angle. GNSS software programs used to output geodetic by default, but nowadays I think grid projection bearings are more common.
?ÿ
Edit to add: if you're adjusting mixed data simultaneously i.e., GNSS and conventional observations, you're going to be using a minimally constrained LS adjustment and relative positional precision to determine whether you meet specs. If you are simply fixing your azimuths at either end and doing a compass rule adjustment, then you're probably going to be looking at closures.
For a long linear traverse like this I would be taking additional GNSS observations throughout the traverse and using LS to adjust.
As mentioned, NAD83(90) coordinates may be measurably different from NAD83(latest).
I just looked at the CCR, and I noticed that the horizontal datum is indeed NAD83(1990). If we used the published coordinates for the initial point (CCR) we're in trouble. Maybe that's why there's such a large difference between 167 and 169, and between 168 and 170.
It's all project-dependent and/or supervisor-dependent.
This project will be used primarily to determine right-of-way lines on both sides of a street. Additionally, we will locate drainage structures, fences, power poles, the edge of pavement, etc. No elevations will be used.
If you are simply fixing your azimuths at either end and doing a compass rule adjustment, then you're probably going to be looking at closures.
That will be about the extent of it.
For a long linear traverse like this I would be taking additional GNSS observations throughout the traverse and using LS to adjust
I'm just guessing, but I'd say the traverse is about a mile long. We made single, 15-minute observations on our beginning and ending pairs of points.
I hope these files will help to clarify what I'm doing. I think the closing linear error is much too big. Any suggestions relating to methods for eliminating this error would be greatly appreciated!
If you are checking a total station traverse against RTK measurements, you need a Combined Factor applied one way or the other so that both are at grid or both at ground. Rather than use someone else's factor at one single point, I would average the CF values for my RTK points.
Yes, this may be his primary problem.?ÿ Calculate the CSF, apply the CSF to your traverse, then compare.?ÿ He traversed a distance of 4600+/-, with a linear error of 0.5ish...?ÿ I'm not sure where you're located, but if a CSF hasn't been applied then I think it needs to be applied to compare the GPS and total station measurements.
Would my linear error of closure be 0.559' (167-169), or 0.495' (168-170)?
Is the total station measurement shorter, or is the GPS distance shorter??ÿ Your final check is the last GPS point in the line.?ÿ If you have a pair of points, and you set on the first GPS point you get to, and you shoot the foresight (the second GPS point) then the foresight is the point you hold fixed, and the bearing to that final foresight is the bearing you are comparing.?ÿ The linear error is the distance error between the GPS and the total station point.
I find it helpful to draw a sketch.?ÿ "A picture is worth a thousand words" and all that.?ÿ I would label the GPS points on the sketch, show the traverse line, any side checks, etc.?ÿ Personally, that's what I would do.
Calculate the CSF, apply the CSF to your traverse, then compare.?ÿ He traversed a distance of 4600+/-, with a linear error of 0.5ish...?ÿ I'm not sure where you're located, but if a CSF hasn't been applied then I think it needs to be applied to compare the GPS and total station measurements.
The OP is located in Orlando, FL.
CSF for FLWE is 0.99994275 (nearest CORS)
That would account for linear difference of?ÿ 0.26ft?ÿ
he OP is located in Orlando, FL.
CSF for FLWE is 0.99994275 (nearest CORS)
That would account for linear difference of?ÿ 0.26ft?ÿ
I was not aware of where he is located, but that's significant.?ÿ
that's significant.
But not enough. It only accounts for half of his error.?ÿ
I believe the rest of the error may exist in the Network VRS solution,?ÿ which I would not rely upon for control. His Hiper SR reads Navstar and Glonass only, no L2C or L5.
that's significant.
But not enough. It only accounts for half of his error.?ÿ
I believe the rest of the error may exist in the Network VRS solution,?ÿ which I would not rely upon for control. His Hiper SR reads Navstar and Glonass only, no L2C or L5.
It sounded like independent (single-point) static sessions of 15 minutes were run. Unless they did a network RTK observation of 15 minutes?
Is the local RTN down there really that poor?
If one adjusts for grid vs. ground and removes half of that error, the raw horizontal misclosure goes from half a foot to a quarter foot. Over the ~4500' distance, that's about a 1:18000 ratio. Depending on equipment used, I wouldn't call that terrible. Looks like about 12 stations, about a couple hundredths to adjust per station?
@rover83?ÿ
From OP original post;
"We started out using 2 RTK-observed points. The first one, point 150, being a 1/4 corner. We closed on 2 RTK-observed points".
He works for the county.
@rover83?ÿ
From OP original post;
"We started out using 2 RTK-observed points. The first one, point 150, being a 1/4 corner. We closed on 2 RTK-observed points".
He works for the county.
Missed the RTK bit...but why in the world run 15 minute RTK observations? Could have tagged several shorter observations at multiple points along the corridor during that time frame and tightened up the whole network.
@rover83?ÿ
Guess they think longer RTK observations get better results. Not sure why they continue to use the board as a free form of education. I have tried to provide training to FieldDog and the county for the last 4 years to no avail. Odd that they can find money to purchase equipment but not training. Yet waste money on doing the job over due to lack of training or education. Even more puzzling that they admit all of this on the forum. This makes it very difficult to help those that fail to help themselves. I'm not bashing FieldDog, this is on managment.?ÿ
The first one, point 150, being a 1/4 corner.
That is not correct. Please see the 8873 Traverse Courses file. My initial post was incorrect because I didn't have the field notes to reference.
CSF for FLWE is 0.99994275 (nearest CORS)
I'm confused about the nearest CORS reference. We use L-Net (Lengemann Corporation) as our network.
?ÿ
Unless they did a network RTK observation of 15 minutes?
That's what we did.
You could apply a CSF to the entire project, or CSFs to individual stations/observations.
Okay, I understand. Should I also subtract 1.6" (0?ø00'19"/12) from each angle too, and then do a?ÿcompass rule adjustment? We don't use least squares.