> StarNet thinks the problem is with your zenith angles (could be an HR bust too) in 3 to 4 or 4 to 5 or possibly backwards...
>
Hey Dave:
No Network Plot with that data? What's up with that, lol?
On a more serious note: How did you get to a closure of more than 1:25000 from Yswami's 1:11000 (and change)? Was it by taking some bad data out of the analysis?
Blunder Detection--just checked field notes
Aloha, Kent and Dave:
I have something funny happening on collection of point #5.
I have a note in my field book about the particular observation. I deleted the point because "rod was not set correct." The recollected it again. Raw data in data collector shows the first collected #5 was deleted. Perhaps, I still messed
There was a chainlink fence in between #28 and #29 that also received asterisks!
I learned two important lessons...
1. The importance of field notes
2. Proper use of LSA to pin point the error (still need to learn more on this!)
The solution seems obvious I could redo #5 and the shot through the chain link fence--which I think I can remove a segment of the fence.
One question regarding fixing this--with assumption redoing is the solution--if I do redo these blunders wouldn't this effect the remaining data previously collected based on this erroneous point?
Thank you so much Kent and Dave!
>
> The Zenith Angles with an asterisk at the end of the line and a StdRes (standardized residual) much larger than 3.0 are obvious candidates for rechecking.
>
Star*Net Blunder Detection
> Your result is similar to Dave's also except you have few extra points that have its zenith angle out of tolerance.
>
> It didn't past Chi-Square test because I don't have enough redundant shots right?
No, the problem is that there appears to be a blunder in one or more of the zenith angles. Possibly it is an incorrect HT (Target Height) value.
One apparent blunder appears here:
[pre]
DV 4-3 564.7711 90-40-06.74 5.469/5.000 'CP3
DV 4-3 564.7807 90-37-34.48 5.469/5.000 'CP3
[/pre]
That's a difference of more than 02'30' in repeat zenith angles between two points 564 ft. apart.
I ran your data as Dave Karoly posted it, but without really noticing that some of the lines are exact repeats. Apparently that's something that the converter created.
I'll have to read Dave's later posts to see if the .DAT file got cleaned up a bit to remove the duplicate entries for the same observations.
Blunder Detection--just checked field notes
> One question regarding fixing this--with assumption redoing is the solution--if I do redo these blunders wouldn't this effect the remaining data previously collected based on this erroneous point?
If you discover that the prior measurements were blunders, just substitute the new measurements and remove the blundered measurements from the adjustment. There's no particular need to repeat measurements that otherwise appear to be without blunders.
Kent caught a mis-entry in one of the backsights (DV).
I fixed that in my data, maybe that is why, I'm not exactly sure.
Star*Net Blunder Detection
Chi square is a test of your assumptions about the data.
If you tell it your centering is 0.005 and it's really 0.05 then you will fail. If you change the centering to 0.05 then it should pass Chi Square. You can make a bad survey pass Chi Square by changing the standard errors you put in the project options.
Blunder Detection--just checked field notes
Aloha, Kent:
Sorry for nagging you again with another question...
If I understand you correctly... I could occupy #4, BS #3 then FS #5. Use that data to replace what appears to a blundered data and that should be sufficient?
My question...isn't point #6 onward are collected based on the error in the blundered #5?
Thanks again!
> > One question regarding fixing this--with assumption redoing is the solution--if I do redo these blunders wouldn't this effect the remaining data previously collected based on this erroneous point?
>
> If you discover that the prior measurements were blunders, just substitute the new measurements and remove the blundered measurements from the adjustment. There's no particular need to repeat measurements that otherwise appear to be without blunders.
Star*Net Blunder Detection
Thanks Dave!
> Chi square is a test of your assumptions about the data.
>
> If you tell it your centering is 0.005 and it's really 0.05 then you will fail. If you change the centering to 0.05 then it should pass Chi Square. You can make a bad survey pass Chi Square by changing the standard errors you put in the project options.
Blunder Detection--just checked field notes
> If I understand you correctly... I could occupy #4, BS #3 then FS #5. Use that data to replace what appears to a blundered data and that should be sufficient?
>
> My question...isn't point #6 onward are collected based on the error in the blundered #5?
Well, the first question I have is whether the problem at Pt.4 was a wrong HI entry. I'll play around with that and see.
It may be just a matter of making some repeat measurements to confirm the HI correction at Pt.4
Blunder Detection--just checked field notes
Aloha, Kent: Thanks!
I can probably do few repeat measurement to verify if Pt. 4 HI is correct tomorrow afternoon. I will remeasure PT. 4 by occupying PT. 3. Then continue with other redundant shots as well.
>
> Well, the first question I have is whether the problem at Pt.4 was a wrong HI entry. I'll play around with that and see.
>
> It may be just a matter of making some repeat measurements to confirm the HI correction at Pt.4
One thing I just noticed...
Your backsights from 4 to 3 zenith angles mis-match by quite a bit...
[pre]
#DV 4-3 564.7711 90-40-06.74 5.469/5.000 'CP3
#DV 4-3 564.7807 90-37-34.48 5.469/5.000 'CP3
[/pre]
There may be a problem with 3 to 4 because commenting the backsights out doesn't make it pass. I would put the sets from 3 to 4, 4 to 3, 4 to 5 and 5 to 4 into a spreadsheet and calculate the individual HD, VD and Elev differences. There may be a problem lurking there.
...
It exceeds the lower bound of Chi Square as a 2D adjustment:
[pre]
Adjustment Statistical Summary
==============================
Iterations = 2
Number of Stations = 33
Number of Observations = 88
Number of Unknowns = 64
Number of Redundant Obs = 24
Observation Count Sum Squares Error
of StdRes Factor
Angles 34 5.657 0.781
Distances 53 1.558 0.328
Az/Bearings 1 0.000 0.000
Total 88 7.216 0.548
Warning: The Chi-Square Test at 5.00% Level Exceeded Lower Bound
Lower/Upper Bounds (0.719/1.281)
[/pre]
The traverse closure is the same:
[pre]
Traverse Closures of Unadjusted Observations
============================================
(Beginning and Ending on Adjusted Stations)
TRAVERSE 1
Error Angular = -94.94 Sec, 32 Angles, -2.97 Sec/Angle
Error Linear = 0.0044 S, 0.3524 E
Horiz Precision = 0.3525 Error in 8989.0594, 1:25503, 39.21 PPM
[/pre]
I have a feeling your horizontal close is about as good as it will get since you have a lot of short courses in there and the geometry isn't the best. If you could do some cross ties that might help. I think the problem is mainly in the vertical.
Apparent HT and HI blunders
Part of the problem was this bit in the input file that I've commented out with the # character:
[pre]
DV 28-27 104.6271 90-39-27.73 5.745/5.000 'CP26
#M 28-27-29 202-10-22.54 185.1539 88-04-43.49 5.745/5.340 'CP28
DV 28-27 104.6285 90-39-28.38 5.745/5.000 'CP26
M 28-27-29 202-10-22.54 185.1539 88-04-43.49 5.745/5.000 'CP28
[/pre]
If you can confirm that 5.340 was an incorrect Target Height, then that is a simple enough fix.
The other part of the zenith angle problem appears to me to be that the Height of the Instrument at Pt.4 was actually 5.069 ft., rather than 5.469 ft.
Making those two changes removes the zenith angle blunders that show up in the adjustment.
After HT and HI corrections, things look good
Okay, here's what the adjustment looks like after making the HT and HT changes I described above and removing one of each identical twin lines from the pairs that were somehow introduced into the .DAT file
[pre]
Statistical Summary
Observation Count Error Factor
Angles 34 1.145
Distances 54 0.854
Az/Bearings 1 0.000
Zeniths 54 1.161
Total 143 1.047
Chi-Square Test at 5.00% Level Passed
Lower/Upper Bounds (0.798/1.201)
[/pre]
The adjustment passes the Chi-Square test and neither the horizontal angles, distances nor the zenith angles show any obvious blunders:
[pre]
Adjusted Zenith Observations (DMS)
From To Zenith Residual StdErr StdRes
2 1 95-52-48.19 -0-00-09.75 10.55 0.9
2 3 89-00-47.20 0-00-06.79 10.24 0.7
3 2 90-53-28.87 -0-00-06.02 10.24 0.6
3 4 89-20-41.02 0-00-13.71 10.12 1.4
4 3 90-37-30.20 -0-00-04.28 10.12 0.4
4 5 88-30-32.35 0-00-02.82 10.28 0.3
5 4 91-30-54.42 -0-00-05.06 10.28 0.5
5 6 89-04-40.81 0-00-05.37 10.52 0.5
5 4 91-30-54.42 -0-00-04.08 10.28 0.4
5 7 87-13-53.32 0-00-07.05 10.16 0.7
6 8 86-40-14.17 0-00-05.15 10.12 0.5
8 6 93-20-30.38 -0-00-04.59 10.12 0.5
8 9 86-32-43.97 0-00-00.04 10.47 0.0
9 8 93-32-42.45 -0-00-05.30 10.47 0.5
9 10 92-18-59.01 0-00-03.91 10.22 0.4
10 9 87-39-10.83 -0-00-04.04 10.22 0.4
10 11 91-04-57.55 0-00-02.84 10.16 0.3
10 7 91-53-43.95 -0-00-05.45 10.25 0.5
11 10 89-02-49.79 -0-00-08.22 10.16 0.8
13 11 83-05-31.81 -0-00-00.75 11.78 0.1
11 10 89-00-58.77 -0-00-04.56 10.16 0.4
11 13 96-54-42.49 0-00-05.09 11.78 0.4
13 11 83-03-27.91 0-00-00.04 11.78 0.0
13 14 90-03-46.35 0-00-09.87 10.39 1.0
14 13 90-05-47.99 -0-00-00.57 10.39 0.1
14 15 88-10-10.90 0-00-09.27 10.69 0.9
15 14 91-56-02.01 0-00-01.04 10.69 0.1
15 16 88-58-41.91 0-00-06.63 10.26 0.6
16 15 90-56-42.85 -0-00-05.74 10.26 0.6
16 17 91-43-58.88 0-00-01.37 10.20 0.1
17 16 88-16-18.42 -0-00-12.96 10.20 1.3
17 18 92-47-14.60 0-00-01.49 10.33 0.1
18 17 87-10-17.13 -0-00-09.79 10.33 0.9
18 19 90-02-50.49 0-00-10.21 10.41 1.0
19 20 89-39-34.00 0-00-06.57 11.33 0.6
20 21 89-52-14.92 0-00-10.66 10.37 1.0
21 22 91-46-21.38 0-00-07.55 10.88 0.7
22 23 91-59-16.71 0-00-00.65 10.61 0.1
23 22 87-39-26.13 -0-00-08.02 10.61 0.8
23 24 92-33-53.01 0-00-06.60 11.34 0.6
24 25 90-22-35.85 0-00-05.45 14.75 0.4
25 26 88-55-24.78 0-00-07.52 10.89 0.7
26 27 87-55-19.68 0-00-06.61 11.33 0.6
27 28 89-26-42.10 -0-00-11.48 13.04 0.9
28 27 90-39-19.55 -0-00-08.18 13.04 0.6
28 27 90-39-19.55 -0-00-08.83 13.04 0.7
28 29 88-04-50.57 0-00-07.08 11.06 0.6
29 30 86-47-53.45 0-00-05.47 13.85 0.4
30 31 88-52-34.09 0-00-05.72 12.42 0.5
31 32 90-36-06.40 0-00-05.70 12.37 0.5
32 33 93-22-22.09 0-00-09.07 10.52 0.9
33 34 93-08-34.79 0-00-15.71 10.15 1.5
34 1 89-03-36.56 0-00-11.30 10.33 1.1
1 2 83-57-00.33 -0-00-00.77 10.55 0.1
[/pre]
Apparent HT and HI blunders
Aloha, Kent:
I can confirm that HI for Pt. 4 is 5.068 ft. per my field note book. However, in my DC raw data I don't have the changes...something happened there.
Unfortunately, I DIDN'T write down the target or the back sight rod height in my field book!!!:-( 🙁 I do remember I had to re-setup again due to chain link fence. It appears like somehow in the process resetting the station my data collector didn't record the changes. It is similar to #4 where I made a mistake and redid the shots but HI was not changed. I must be doing a mistake that causing this.
You commenting out the first line is correct. Logically, I redid it because I was happy with first sets. Target Height got to be 5.0 ft.
Thanks Kent...must be pretty late now in Texas!! Hopefully, I am not upsetting Mrs. McMillan for keeping you at the computer! 🙂
> Part of the problem was this bit in the input file that I've commented out with the # character:
>
> [pre]
> DV 28-27 104.6271 90-39-27.73 5.745/5.000 'CP26
> #M 28-27-29 202-10-22.54 185.1539 88-04-43.49 5.745/5.340 'CP28
> DV 28-27 104.6285 90-39-28.38 5.745/5.000 'CP26
> M 28-27-29 202-10-22.54 185.1539 88-04-43.49 5.745/5.000 'CP28
> [/pre]
>
> If you can confirm that 5.340 was an incorrect Target Height, then that is a simple enough fix.
>
> The other part of the zenith angle problem appears to me to be that the Height of the Instrument at Pt.4 was actually 5.069 ft., rather than 5.469 ft.
>
> Making those two changes removes the zenith angle blunders that show up in the adjustment.
Back in the day (pre-Star Net), if I found that I had an angular blunder, I most-times would be able to find out where it occurred thus:
Using the unadjusted angles, run unadjusted traverses each direction. Where the coordinates came together, most likely the source of the bust.
Question for yswami
One question that I meant to ask was what flavor of data collection software you're using. What I concluded (possibly incorrectly) from your recent work was that at one of the stations where the Instrument Height was entered with a value 0.40 ft. too high, the data collector failed to flag a gross discrepancy when you got a shot to the backsight. Since both the instrument station and the backsight station had elevations that should have already been computed by the DC, the discrepancy should have been simple enough to identify on the spot.
Does your DC software not do this? If not, which software is the culprit?
Question for yswami
Aloha, Kent:
I use Survey Pro (TDS--currently is know Spectra Precision aka Trimble). Version 5.2
I use the automatic repetition feature where the data collector software will collect specified number of sets in the order set in the preferences. It does alert me when horizontal angle tolerance are higher and normally need to redone. I will post the settings later today (after 12 pm Hawaii time).
Honestly, I don't know what happened here. I would like to think that I am the cause of the blunder in the way I handled the software when I decided to redo the point. In both cases where the blunders were discovered by you and Dave, I had point deleted and reshot. However, I will try to recreate the situation again and see if I can repeat it. If it is a software issue I will send the developer a note.
Thank you Kent. Will be back at my computer around noon.
> One question that I meant to ask was what flavor of data collection software you're using. What I concluded (possibly incorrectly) from your recent work was that at one of the stations where the Instrument Height was entered with a value 0.40 ft. too high, the data collector failed to flag a gross discrepancy when you got a shot to the backsight. Since both the instrument station and the backsight station had elevations that should have already been computed by the DC, the discrepancy should have been simple enough to identify on the spot.
>
> Does your DC software not do this? If not, which software is the culprit?
Question for yswami
Yswami,
I use the same Survey Pro software. I think what Kent is referring to is your back sight check routine when your solving your occupy/back sight setup. When you set up on a station and enter your HI and BS HR and shoot your back sight, the software will give you the vertical and horizontal discrepancy between what was measured and what was computed using the numbers that you entered. It's a routine check that would reveal the discrepancy that went undetected. If you entered a bad BS HR or instrument HI, the discrepancy would show up there when you sighted and shot your BS. Careful attention to those checks will save you a lot of grief later on.
Just because I'm paranoid, doesn't mean they aren't out to get me.
> Back in the day (pre-Star Net), if I found that I had an angular blunder, I most-times would be able to find out where it occurred thus:
Draw a perpendicular to the closure leg (that is, the little 0.3' misclosure leg) through the closing point. It should point to the station with the busted angle. That is if the bust is angular, and it is isolated to one station.
Question for yswami
> I use the same Survey Pro software.
Just out of curiosity, how does Survey Pro handle sideshot measurements to existing stations? That is if you position, say Pt.49 from one control point and then get a shot to it from another control point, do you have to assign a new point no. when the second tie is logged?
One odd feature of the data entry file was that quite a few of the stations had i.d. nos. in the descriptors that didn't match the pt. i.d. nos used by the DC and I had to wonder why that would be.