Aloha Friday everyone!
After my first attempt to create control network I tried it again.
Few things I did differently from my first attempt...based on wonderful guidance and advice provided previously.
1. I used forced centering method with tripods and tribrachs. (NO Prism poles)
2. Use Seco regular prisms instead of 360º mini prisms.
3. Simplified the network. Eliminated short traverses and questionable line of sight that I can fix i.e. chain link fences etc.
4. Multiple redundant shots.
5. I was very careful when I measure up at instrument and also at the BS and FS prisms. Measure twice and write once.
6. Since this will be my one and only network I also did it with the "slow method." I used the instrument it self at each station to center and level the tribrachs to reduce the centering error.
7. I turned 3 sets of angles with distance.
The disappointing part there is still a serious blunder in my work. Horizontal angles seems to be tight; however, the vertical is not good. Not sure how to figure out where the blunder occurred and how to rectify? Will I be able to fix this without starting all over again? I know it depends…:-)
Carlson Survey’s SurvNet LSA program gave the following result
After maximum of 10 Iteration solution did not converge within 0.002 with confidence interval of 95%
Closure before adjustment
Total angular error: 000-00’ 09.5"
Distance traversed 8176.828 Closure 1:824040
Elevation error: 0.297
I looked through the raw data but I don’t think I have enough experience to understand them completely. ServNet for an example shows some sets have “Vertical Distance from xx point to xx point exceeds tolerance.” Difference up to 0.6ft are shown in the report but I can see the outlier here. However, I cannot see where the outlier is in the raw data—if it is an outlier.
Thought if someone could take a look at the data might be able to help to answer my questions.
I use Survey Pro, I can email the TDS raw data file if someone be willing to look at it. Last time Dave kindly converted them to StarNet readable format. This allowed Kent, Dave and other to point out the blunders.
Any help and guidance will be very, very much appreciated. Mahalo!
If this is a traverse through the same points we aught to have a look at both sets of data. Please email the .raw files to me via the profile address.
Aloha, Norman: Just emailed you a message. Didn't know how to attach a file to it. Have to wait for your reply.
Thank you so much!
Check your email!
Aloha,
Just sent it to you. Thank you:-)
I would highly recommend using a field book on the first go round.
Manually check and calc. the data yourself.
Have someone check your notes.
This way you have a way of seeing HOW a software program is using your data.
Jumping into things directly by utilizing robotic routines and least square adjustments is jumping the gun in my opinion.
Averaging vertical (zenith) angles was discussed recently as well as how to calculate the vertical difference between stations. You just have to figure out how to extract that data from your raw data file.
Forward readings: I would calc. the vertical diff. 3 times, then average the 3 vertical diff. values.
Reverse readings: Do the same.
Then average the final forward vertical diff. with the final reverse vertical diff. This will be the number you will use to calculate your elevations.
THAT is how you detect blunders...by working with your data. It will become very apparent where the blunders are (if any). Could very well just be the format that it is being fed into the machine...
EDIT: It seems you probably have plenty of redundant measurements for blunder detection...you just have to reduce your notes!
OK. It's not converging for me. I see a couple of things in there that may be causing it. Some mis-entered measure ups (When you aren't closing vertically 99% of the time its a measure up error).
SO what we will do the data in pieces. The first few setups, then add a few more, and a few more until we identify the sticking point.
# Input Field File : U:Monastery6-05-15a MONASTERY CONTROL.RAW
# Date Processed : 06-05-2015 17:12:26
.Units FeetUS
.Units DMS
.Order AtFromTo
.Sep -
.Delta Off
.3D
C 1 -992.63700 -14640.18700 412.43300 ! ! ! 'CP1
C 2 -748.43300 -14553.92400 439.46700 ! ! ! 'CP2
# Job : 5-17-15 MONASTERY CONTROL
# Date : 05-17-2015
# Time : 11:12:20
DV 2-1 260.4578 96-02-37.25 5.449/4.995 'CP1
M 2-1-3 178-15-18.29 393.1249 89-12-04.34 5.449/5.435 'CP3
M 2-1-3 178-15-17.48 393.1241 89-12-05.14 5.449/5.435 'CP3
M 2-1-3 178-15-16.18 393.1224 89-12-02.72 5.449/5.435 'CP3
M 2-1-3 178-15-17.32 393.1238 89-12-04.07 5.449/5.435 'CP3
DV 3-2 393.1337 90-48-58.03 5.374/5.300 'CP2
M 3-2-4 180-26-30.51 564.7913 89-21-17.89 5.389/5.615 'CP4
M 3-2-4 180-26-32.62 564.7921 89-21-17.89 5.389/5.615 'CP4
M 3-2-4 180-26-33.43 564.7944 89-21-16.92 5.389/5.615 'CP4
M 3-2-4 180-26-32.19 564.7926 89-21-17.57 5.389/5.615 'CP4
DV 4-3 564.7946 90-39-29.74 5.580/5.235 'CP3
M 4-3-5 192-57-19.64 364.6901 88-31-44.55 5.580/5.365 'CP5
M 4-3-5 192-57-22.07 364.6925 88-31-43.09 5.580/5.365 'CP5
M 4-3-5 192-57-19.80 364.6901 88-31-43.41 5.580/5.365 'CP5
M 4-3-5 192-57-20.50 364.6909 88-31-43.68 5.580/5.365 'CP5
DV 5-4 364.6979 91-29-18.96 5.319/5.415 'CP4
M 5-4-6 150-25-58.87 268.5700 89-00-17.85 5.319/5.630 'CP6
M 5-4-6 150-25-58.39 268.5696 89-00-18.18 5.319/5.630 'CP6
M 5-4-6 150-25-58.88 268.5693 89-00-17.37 5.319/5.630 'CP6
M 5-4-6 150-25-58.71 268.5696 89-00-17.80 5.319/5.630 'CP6
DV 6-5 268.5818 91-01-03.79 5.570/5.170 'CP5
M 6-5-7 198-18-52.06 552.6290 86-39-46.40 5.570/5.613 'CP7
M 6-5-7 198-18-52.22 552.6272 86-39-46.40 5.570/5.613 'CP7
M 6-5-7 198-18-52.22 552.6265 86-39-45.92 5.570/5.613 'CP7
M 6-5-7 198-18-52.16 552.6276 86-39-46.24 5.570/5.613 'CP7
DV 7-6 552.6418 93-21-04.46 5.565/5.420 'CP6
M 7-6-8 260-50-47.58 253.7382 88-49-45.89 5.565/5.790 'CP8
M 7-6-8 260-50-47.09 253.7358 88-49-46.22 5.565/5.790 'CP8
M 7-6-8 260-50-47.09 253.7371 88-49-45.89 5.565/5.790 'CP8
M 7-6-8 260-50-47.26 253.7370 88-49-46.00 5.565/5.790 'CP8
DV 8-7 253.7464 91-11-40.45 5.745/5.413 'CP7
M 8-7-9 270-35-24.79 415.6023 92-23-40.35 5.745/5.890 'CP9
M 8-7-9 270-35-24.46 415.6012 92-23-40.83 5.745/5.890 'CP9
M 8-7-9 270-35-24.46 415.6007 92-23-40.66 5.745/5.890 'CP9
M 8-7-9 270-35-24.58 415.6014 92-23-40.61 5.745/5.890 'CP9
DV 9-8 415.6220 87-31-18.98 4.640/5.630 'CP8
DV 9-8 415.6207 87-31-20.28 5.149/5.630 'CP8
M 9-8-10 50-17-26.67 302.0197 88-53-43.87 5.149/5.535 'CP10
M 9-8-10 50-17-26.82 302.0187 88-53-44.36 5.149/5.535 'CP10
M 9-8-10 50-17-26.34 302.0179 88-53-43.87 5.149/5.535 'CP10
M 9-8-10 50-17-26.61 302.0188 88-53-44.03 5.149/5.535 'CP10
DV 10-9 302.0289 91-07-22.87 5.479/5.005 'CP9
M 10-9-11 329-42-12.70 545.2189 91-59-33.36 5.479/5.595 'CP11
...
It doesn't like the very first shot (2-1) for zenith angle, so we will free up the elevation on point 1.
At 9 backsighting 8 you have a measure up bust that you corrected in the field but it's still in the data
M 8-7-9 270-35-24.79 415.6023 92-23-40.35 5.745/5.890 'CP9
M 8-7-9 270-35-24.46 415.6012 92-23-40.83 5.745/5.890 'CP9
M 8-7-9 270-35-24.46 415.6007 92-23-40.66 5.745/5.890 'CP9
M 8-7-9 270-35-24.58 415.6014 92-23-40.61 5.745/5.890 'CP9
#DV 9-8 415.6220 87-31-18.98 4.640/5.630 'CP8
DV 9-8 415.6207 87-31-20.28 5.149/5.630 'CP8
M 9-8-10 50-17-26.67 302.0197 88-53-43.87 5.149/5.535 'CP10
M 9-8-10 50-17-26.82 302.0187 88-53-44.36 5.149/5.535 'CP10
M 9-8-10 50-17-26.34 302.0179 88-53-43.87 5.149/5.535 'CP10
M 9-8-10 50-17-26.61 302.0188 88-53-44.03 5.149/5.535 'CP10
Remark that line out with a "#"
> I would highly recommend using a field book on the first go round.
> Manually check and calc. the data yourself.
> Have someone check your notes.
>
> This way you have a way of seeing HOW a software program is using your data.
>
> Jumping into things directly by utilizing robotic routines and least square adjustments is jumping the gun in my opinion.
>
>
>
> Averaging vertical (zenith) angles was discussed recently as well as how to calculate the vertical difference between stations. You just have to figure out how to extract that data from your raw data file.
>
> Forward readings: I would calc. the vertical diff. 3 times, then average the 3 vertical diff. values.
>
> Reverse readings: Do the same.
>
> Then average the final forward vertical diff. with the final reverse vertical diff. This will be the number you will use to calculate your elevations.
>
> THAT is how you detect blunders...by working with your data. It will become very apparent where the blunders are (if any). Could very well just be the format that it is being fed into the machine...
>
> EDIT: It seems you probably have plenty of redundant measurements for blunder detection...you just have to reduce your notes!
Aloha, Imaudigger:
I completely agree with you on all points! I just wished I had a bit more exposure and mentoring doing the things manually. Always felt guilty about pushing the button! Because I don't know what is happening behind it. I am studying a little bit on the side to understand things better but still very, very long ways to go!
You have inspired me to pursue this deeper than I am doing.
I have several redundant shots. Reducing the note is my other challenge. I noticed Norman posted the data nice and clean but when I look both in Survey Pro and SurvNet it is a bit confusing!
> At 9 backsighting 8 you have a measure up bust that you corrected in the field but it's still in the data
>
>
> M 8-7-9 270-35-24.79 415.6023 92-23-40.35 5.745/5.890 'CP9
> M 8-7-9 270-35-24.46 415.6012 92-23-40.83 5.745/5.890 'CP9
> M 8-7-9 270-35-24.46 415.6007 92-23-40.66 5.745/5.890 'CP9
> M 8-7-9 270-35-24.58 415.6014 92-23-40.61 5.745/5.890 'CP9
>
> #DV 9-8 415.6220 87-31-18.98 4.640/5.630 'CP8
> DV 9-8 415.6207 87-31-20.28 5.149/5.630 'CP8
> M 9-8-10 50-17-26.67 302.0197 88-53-43.87 5.149/5.535 'CP10
> M 9-8-10 50-17-26.82 302.0187 88-53-44.36 5.149/5.535 'CP10
> M 9-8-10 50-17-26.34 302.0179 88-53-43.87 5.149/5.535 'CP10
> M 9-8-10 50-17-26.61 302.0188 88-53-44.03 5.149/5.535 'CP10
>
>
> Remark that line out with a "#"
Norman:
Are you using a converter to get Yswami's data from SurvPro into Starnet? Starnet seems to format the data so that it's just plain easier to scan and pick out problems.
After several such deletions (no more than my field crews typically have) I've got a pretty good solution. One issue remains with zeniths at inst Pt 15, backsighting 14. Probably a measure up issue. You have HI=5.299 and Ht of the backsight at 5.17 (first entered as 5.795, then changed). Please check your notes and see if that is correct. 5.22 on the backsight would work better.
> Are you using a converter ..?
Yes. One of the side benefits of StarNet is that the data format is easy to read.
> At 9 backsighting 8 you have a measure up bust that you corrected in the field but it's still in the data
>
>
> M 8-7-9 270-35-24.79 415.6023 92-23-40.35 5.745/5.890 'CP9
> M 8-7-9 270-35-24.46 415.6012 92-23-40.83 5.745/5.890 'CP9
> M 8-7-9 270-35-24.46 415.6007 92-23-40.66 5.745/5.890 'CP9
> M 8-7-9 270-35-24.58 415.6014 92-23-40.61 5.745/5.890 'CP9
>
> #DV 9-8 415.6220 87-31-18.98 4.640/5.630 'CP8
> DV 9-8 415.6207 87-31-20.28 5.149/5.630 'CP8
> M 9-8-10 50-17-26.67 302.0197 88-53-43.87 5.149/5.535 'CP10
> M 9-8-10 50-17-26.82 302.0187 88-53-44.36 5.149/5.535 'CP10
> M 9-8-10 50-17-26.34 302.0179 88-53-43.87 5.149/5.535 'CP10
> M 9-8-10 50-17-26.61 302.0188 88-53-44.03 5.149/5.535 'CP10
>
>
> Remark that line out with a "#"
Aloha, Norman:
Thank you for posting the data!
How did you get you data so clean. I am seeing very confusing raw data in SurvNet. See attached screen shot
It is interesting that the changes in the field still used by office software.
SurvNet show vertical distancel error of 0.551 from point 8 to 9.
I also notice on SurvNet error report measurement from 13 to 14 has too many sets. I cancelled the first three set and redid the shots, but it kept all of them. 
> After several such deletions (no more than my field crews typically have) I've got a pretty good solution. One issue remains with zeniths at inst Pt 15, backsighting 14. Probably a measure up issue. You have HI=5.299 and Ht of the backsight at 5.17 (first entered as 5.795, then changed). Please check your notes and see if that is correct. 5.22 on the backsight would work better.
Aloha, Norman: Verified my notes it is 5.17. The first entry of 5.795 was previous backsighting ht. at Pt 13. Which I forgot to change when I moved the next point.
Thank you
> After several such deletions (no more than my field crews typically have) I've got a pretty good solution. One issue remains with zeniths at inst Pt 15, backsighting 14. Probably a measure up issue. You have HI=5.299 and Ht of the backsight at 5.17 (first entered as 5.795, then changed). Please check your notes and see if that is correct. 5.22 on the backsight would work better.
Aloha, Norman: Found the problem. I converted wrong decimal point when I converted from Slant to true vertical! It should be 5.26ft. Phew! Thanks to field notes! Instead deducting 0.01 I deducted 0.1
> 5.22 on the backsight would work better.
I've snagged your January data off the board and used it to convince myself that this is exactly what happened. Reading the tape as 5.17 when you where looking at 5.22 is a common mistake, easy to make. Anyway, we are down to fussing with 0.05'. I will email you the listing file and edited dat file.
In general the data looks excellent.
> Aloha, Norman:
> Thank you for posting the data!
> How did you get you data so clean. I am seeing very confusing raw data in SurvNet. See attached screen shot
That's the StarNet format. I converted your .raw file and that's what you get.
> It is interesting that the changes in the field still used by office software.
> SurvNet show vertical distancel error of 0.551 from point 8 to 9.
Inspect the .dat file I emailed you. There are a number of line remarked out with a "#". These line are followed by another shot you took with the measure up corrected, which overwrites the point in the dc. Typical field procedure. But the shot with the bad measure up still remains in the raw data and crops up when you re-run the raw data in the office
> I also notice on SurvNet error report measurement from 13 to 14 has too many sets. I cancelled the first three set and redid the shots, but it kept all of them.
At first the StarNet converter errored out and refused to convert you data but it alerted me to the problem lines, which I remarked out of the .raw file. Those must have been your deletions.
If you're using survey pro with your S6, you can accurately measure the slant height to the bottom notch and have the software convert to true HI. That is 1 way to eliminate an error source...
please post the listing file...