California Surveyors Forums

[I dug up the deposition transcript from a fellow that doesn't keep notes. The questioning were focusing on his "field notes" - raw data. In all fairness, the guy had been pummeled with his lack of understanding of land surveying up to this point].

17 Q When you say "remaining data," what are you

18 specifically referring to? Read it into the record.

19 A 90.0711 and 85.945694.

20 Q Okay. And then there's a couple of dashes and

21 it's zero plus 02.

22 A Those are -- that's what I labeled the point.

23 Q Okay. What does the AR stand for? Do you see

24 the AR in that -- about a third of the way into that line?

25 A I don't remember it. I can't remember if it's


148

1 azimuth reading.

2 Q What does the SS stand for?

3 A I don't remember.

4 Q Okay. Well, let's see if we can hit the hat

5 trick here. Then what does the SD stand for?

6 A I don't remember.

7 Q Could that be slope?

8 A I don't remember really.

9 Q Could it be slope distance?

10 A It could be. I don't remember.

11 Q Okay. And the AR you thought maybe is some kind

12 of azimuth reading?

13 A Uh-huh.

14 Q And the SS, you really don't know what that would

15 be?

16 A I don't remember. I think the ZE is the zenith.

17 Q Why would a slope distance be recorded, if you

18 know?

19 A It tells you, I guess, the total distance both

20 vertically and horizontally that the prism is from the

21 total station, I assume.

22 Q Okay. Now, assuming the total station is not at

23 the same elevation as the point that you're measuring,

24 would that line be the slope distance to the prism? Is

25 that what you're talking about?


149

1 A I'm not sure. I don't know for sure.

2 Q Okay. And you think that's not relevant to a

3 coordinate?

4 A It may be. I'm not sure.

5 Q Okay. Now, I'd like you to take a look at 3 --

6 okay. I'd like you to look at 3B.

7 A Can I go back to that at all?

8 Q Absolutely.

9 A So I see looking at other data that I was

10 incorrect in what I thought this was. What line was it?

11 36 we were talking about was --

12 Q Correct.

13 A I don't know what that data is.

14 Q At all? Nothing in that line?

15 A No. I don't remember what they're used for.

16 Q Let me ask you this. Could an AR be angle right,

17 if you know?

18 A I don't know.

19 Q Could ZE be Zenith?

20 A I believe so.

21 Q Could SD be slope distance?

22 A I don't know.

23 Q Could SS mean side shot?

24 A I don't know.

25 Q Okay. Occupied DP mean occupied point or OP mean

150

1 occupied point?

2 A That sounds correct.

3 Q And FP, could that be foresight point?

4 A That sounds correct.

5 Q What is an occupied point?

6 A The point that the prism is over.

7 Q Okay. What is a foresight point?

8 A I don't remember.

9 Q Okay. What is the side shot, if you know?

10 A I don't remember.

11 Q Okay. And the slope distance, you don't know

12 what that is?

13 A No.

[This was a 6+ hour deposition - he knew he was in trouble by now]

RTK/RTN will not yield the accuracy required to certify an ALTA Land Title Survey - even with optimum conditions. One out every 10 measurements will fall well outside of the accuracy standards. The surveyor has no means to determine which one of the ten measurements are bunk. To certify to having met the ALTA "minimum" standards using RTK, especially RTN, is negligence.

As with any measurement technology, proper application is required in order to obtain acceptable results. Accuracy is the product of equipment, procedure and expertise, and that pertains to RTK as well as it does to terrestrial measurements. RTK (and its RTN variant), properly used in the right circumstances, can meet ALTA standards. While the quoted statement is correct in some instances, it is far too broad to be useful.

RTK vector data comes with standard error information if you know where to look for it. By combining redundant weighted vectors acquired under suitably different conditions (particularly constellation configuration) into a least squares adjustment, the probable error of the resulting position can be estimated reliably and compared with the ALTA or other pertinent accuracy specification to determine acceptability.

I'm kind of partial to 3 vectors per point, but I've run with 2 when the vectors are short enough and the correlation tight. At other times I've gone with 4 or more before getting comfortable with the result.

Trying to use RTK to do an ALTA in a highrise block in San Francisco would not be a good idea, in my opinion. But for an ALTA requiring a section breakdown in the Sacramento Valley, it's probably the most efficient and effective tool available in the hands of a knowledgeable user.

Rob, you are permitted to use all of the Curt Brown quotes for the purpose of this presentation. Good luck!

Jim Frame stated "....While the quoted statement is correct in some instances, it is far too broad to be useful.

RTK vector data comes with standard error information if you know where to look for it. By combining redundant weighted vectors acquired under suitably different conditions (particularly constellation configuration) into a least squares adjustment, the probable error of the resulting position can be estimated reliably and compared with the ALTA or other pertinent accuracy specification to determine acceptability".

The horizontal standard error of a GPS RTK unit is 1 cm which is equal to 2 cm (0.06') at 2 sigma. This error of 0.06' does not account for PPM, the bubble, the rod etc. - which will increase the size of the error ellipse. The RTN allows for longer measurements which run up the error due to PPM. Also, this does not take into account the 10% of the measurements that fall outside of the 2 sigma ellipse without any other error sources considered. However, I agree with Jim Frame's assessment - it can be done with multiple measurements which are post processed. The challenge is there are fewer and fewer people that have post processing capabilities.

It has been my experience, the oftentimes the land surveyors:

1. Visit the point once.
2. Collect the same point several times or commit several "measurements" in consecutive order over a few minutes or seconds
(worthless) thinking it is somehow more accurate.
3. Compare two values and mean them - if the values are within 0.04' they believe they have 0.02' accuracy.
4. Sometimes scale the points, sometime not, to "ground" - whatever that means.
5. Do not use (or understand?) the line accuracy/ellipse between two points.

Again, Jim, you are correct in that it can be done. Riddle me this, if you are going to make use of constellation variations and visit the points 3 or 4 times and post process the data - why wouldn't you simple perform one 12-15 minute static session?

I appreciate the finer points you have introduced to the conversation.

DWoolley

If your wife packs you a lunch, eat it, don't under any circumstances should you ever bring it home again.

Happy wife, happy life.

The horizontal standard error of a GPS RTK unit is 1 cm which is equal to 2 cm (0.06') at 2 sigma.

With old equipment and/or network accuracy. With newer receivers and shortish vectors 0.005 m 1-sigma relative accuracy is more like it under good conditions. With multiple observations you can beat that down a bit further.

Riddle me this, if you are going to make use of constellation variations and visit the points 3 or 4 times and post process the data - why wouldn't you simple perform one 12-15 minute static session?

A single fast-static session is almost always better than a single RTK observation, particularly if you use the precise ephemeris. But it doesn't provide enough constellation change for my comfort level, and no check on the iono modeling or tropo delay, so I pretty much never rely on a single fast-static session. Setup time is significantly longer unless you run fast-static with a bipod, which I don't because bipods are prone to small off-plumb movements and much more susceptible to toppling, and I don't want to stand there monkeying with it for 15 minutes. (As a matter of habit, f I'm going to all the trouble of setting up a tripod, I'm likely to stay for at least 20 minutes, and I like 30 even better.) If drive time is an issue I'll consider fast-static or even multi-hour observations, but that's not typical for the kind of boundary work I do.

With easy drive access I can run though a set of control points with RTK pretty quickly, and come back the next day to repeat. I use a 10-minute bubble on the bipod pole, which helps to keep the centering errors low. On both days I'm likely to spend a few hours digging for additional points or picking up features with a total station, then go back and RTK the control points again. It's pretty easy to get two or more observations on each control point that way.

I always post-process RTK. I export a g-file and bring it into Star*Net along with my total station and level data (if applicable). It's just part of my standard workflow. I acknowledge that this isn't typical for the profession, and that a lot of RTK work out there is scarily haphazard.

The questioning were focusing on his "field notes" - raw data.

I wouldn't shy away from the opportunity to be deposed about my raw data -- having written a bunch of software to parse and reformat RW5 files as input data to various applications, I'm pretty familiar with the format. But even though I rely on the RW5 for the bulk of my "field notes," I still tote a book and record details that don't conveniently fit into the 20 or so characters available to me on the data collector. Those notes and the ubiquitous cell-phone photos round out the record of what I find in the field.

It takes 8 years to learn how to do the work...then 8 more years to learn how to not screw it up.

Lee Hixson wrote:It takes 8 years to learn how to do the work...then 8 more years to learn how to not screw it up.

true that

dosen't it only take 6 years experience to get Licensed? maybe things changed
obviously Engineers are much smarter as they can fast track with nearly none

To be a proficient land surveyor, you have to be able to make decisions. Decisions that you are prepared to defend, when questioned. If you can't make decisions, find another line of work.

Always be prepared to think on the fly. No matter how detailed and comprehensive your plan is before you go in the field, it is likely to require alteration when your boots get on the ground. It's just the nature of the beast.

Share your work plan with your staff. This will make them feel engaged, will serve as mentoring, which is being lost, and may just help save your behind if you make an error. If all they know is where you direct them, without knowing the big picture, they learn very little. They may even surprise you with something you had not considered.

I suggest that someone stop using the expression "riddle me this," it sounds like something a junior high schooler would say, no offense to anyone in particular.

I'm not convinced RTK is worse than Faststatic at this point, the jury is still out. RTK has advance quite a bit in the past few years.

I think redundancy is more important. It's looking like 3 (maybe 60 second) RTK shots at different times of day may be better than one half hour Faststatic session. Redundancy is the key. I caught a GNSS issue not because of RTK versus Faststatic but because I had multiple observations. A weird vertical bump crept in at a certain time of day in observations using both methods.

DWoolley wrote:In years past, most data collectors were limited to 18 characters...

Mr. Smith wrote:Think of field notes as a liability reducer...

Thanks for the explanations, always keep learning.

Peter: indeed, 6 years "to get licensed." But "learning how to do the work" takes--at least for me--longer than that. At the 8-year mark I was feeling my oats. Then, in hindsight, throughout the next 8 years I learned from my mistakes what I didn't learn the first time around. My best insight? How easy it is to allow yourself more confidence than you deserve.

Dave: I totally agree about good monument descriptions. If you look at my maps over my 37-year career you'll notice a decided increase in the length of my descriptions. From short one-liners I graduated to 2 or 3 lines squeezed in close to the symbol. After that came 4 or 5-line descriptions, followed by (as needed) short paragraphs off to the side. This parallels the increasing length of my narratives and notes; I want my resolutions to be definitive and crystal clear.

Mr. Karoly:

Not a fan of the original Batman with Adam West and Frank Gorshin as the Riddler?

I probably watched to much Batman as a kid. Fortunately, today I have little kids that now camouflage my regular viewings.

sus deseos son ordenes para mi.

I have tested the GNSS units by recording 1 second measurements over a 24 hour period approximately 70' from the CGPS station and then, turned off the CGPS station causing the unit to bounce to the next closest -
about 10k - unit for another 24 hours of 1 second measurements. No other sources of error introduced - unit to unit. The results? Exactly what the specifications state they will be - nothing more, nothing less. It stands to reason. The manufactor has a vested interest in reporting the best possible accuracy without misrepresentation. I post processed the 24 hours of data to several stations to get the "true" answer. The outliers were 0.3' horizontal and 0.5' vertical - with no rhyme or reason - which occurred over several minutes of measurement. This is the 5%. A user may think the 5% is likely to "be within a tenth", nope. I graphed the results in Excel. RTK and especially, RTN with a remote base unit is a grab bag of accuracy 5% of the measurements.

Anyone else wonder why all manufactor's specifications report standard deviation (68%)? It is the lowest accuracy number i.e 1 cm they can advertise and still he true. I contend few people care about standard deviation. Users in the professional community are generally interested in 2 sigma i.e. 95%.

I am enjoying this thread. Thank you to those that have posted.

DWoolley

Think twice before calling a found monument as being off. Death to the pincushion! Just reviewed a map where they called a monument off by 0.01'. That's some amazing measuring skills.

My understanding is that the presentation is for FS Exam prep students, which I take to be LSIT candidates.
With this in mind, Mr. Wooley's, "A leaning rod is a falling rod" tidbit is perfect, and one I use as well.
In the spirit of considering the direction this thread has gone,

1. Keep it Simple Stupid,
2. Learn to understand where your errors come from,
3. Check your backsite,
4. Redundancy checks, have them built into your workflow and understand them and why you have them,
5. Take the shot, talking about whether you should or not is a waste of time,
6. Learn to book your angles (field notes) and use your instrument without a collector, because collectors fail,
7. Take care of your equipment, it doesn't work you don't work....

Could go on and on....

Well this happens... More so today with GPS.

I got a call today to look at a boundary dispute of one inch. It reflects poorly on the profession.

http://www.amerisurv.com/PDF/TheAmerica ... l10No5.pdf

I have tested the GNSS units by recording 1 second measurements over a 24 hour period

The results aren't surprising, but I'm not sure they're very useful, either. No responsible surveyor is going to hang a position on a single-epoch observation.

Using modern equipment, the odds of a 3-minute RTK observation on an unobstructed point with a short (<5 km) vector having a horizontal error greater than 0.1' without alerting the user to a problem in real time are fairly small. The odds of two such observations taken several hours apart having an error of the same magnitude and direction are very, very small.

I often run through points with both RTK and a total station, and I typically see horizontal RTK residuals on the order of 0.02' or less for very short (<1000') vectors in the open. For some work that's not tight enough, but there's a crossover point in project size beyond which redundant RTK is more accurate than traversing. Throw in its wonderful capability as a search tool, and I can't see any reason not to use it for boundary work.

P.S. I'm laughing a bit as I find myself advocating the benefits of RTK. I was an RTK skeptic for many years, having seen lots of the problems that arose from the pie-in-the-sky sales job that the equipment vendors perpetrated in the early years of the technology. (Some sales people, and more than a few users, remain appallingly unfamiliar with the limits of the technology, as Dave has alluded.) I've only been using RTK regularly for about 3 years, and I was very conservative in introducing it into my workflow. But I find it to be an extremely valuable measurement tool when used with care.

Jim Frame wrote:....The results aren't surprising, but I'm not sure they're very useful, either. No responsible surveyor is going to hang a position on a single-epoch observation.

Actually, if the user takes several measurements (epochs) over a span of 3-4 minutes, in the unobstructed conditions Jim Frame has described, the coordinate is unlikely to measurably vary from a single epoch. In other words, there is little to be gained by collecting several epochs over a few minutes - only that the measurements are repeating themselves (even those out 0.3 from the true answer). I smile when I see someone standing in an intersection plucking away at the registration button on their controller.

When the RTK/RTN solutions are plotted over a timeline they resemble long lazy waves - as opposed to EKG lines bouncing over the baseline with each beat of the heart. Capturing a single epoch anytime over a 5 or 10 minute span - although not recommended for other reasons - will yield the same quality solution as collecting and meaning several measurements over the same period of time.

I found the fact that the northing and easting are uncorrelated to be unsettling. Restated, the "waves" of northing and easting have no relation to each other. As the easting moves toward the true answer the northing may very well be moving away from the true answer. This means the coordinates point to point over a period of time are not relative. The value of the northing - close to the true value - and a wayward easting could very well be the opposite 30-40 minutes down the road at the next point i.e good easting, bad northing. This plays heck with the inverse for a bearing and/or distance for accurate mapping. The 30-60 minutes between points can be realistic if we consider loading, unloading, setting out cones, taking a call, getting a description etc. Staking point to point over 25' is less pronounced, but the vertical can vary greatly throughout a few hours. If dealing with a 6% grade the project can absorb the vertical variation. A 1-2% grade, not so much. Add PPM and obstruction, ridiculous.

If I can find and screen capture the graphic results of my test I will post them.

Good conversation, Jim Frame, thank you. I think we're not in disagreement.

DWoolley

Measure twice - cut once.

A proper survey is an accurate survey, which is not necessarily the result of a precise survey.

A wise old surveyor told me, If you're searching for a monument and don't find it, you probably looking in the wrong spot. Check your calcs, check your field measurements.

The same surveyor, gave me my first field book, handed me a pencil, then jerked it back, bit the eraser off then handed it to me. He said "never erase in your field book, just strike a line through it.

If you use a schonstedt metal detector, (pipe finder, wiz banger what ever nick name got for it), to locate a monument, dig it up and verify it. Never assume it's the monument.

Porcupine corner "Pin Cushion" what corner do we tie? L.S. all'em !@#$%...:)

Thank you all for the great tips and quotes, and particularly the "hijack" (tangential detour?) on the topic of RTK! Many posters take offense to hijacks, but I think that was an opportune deviation from the OP to a topic that far too few of us understand at the depth of some of our contributing colleagues. Another reason for people to attend the CLSA Conference! Always plenty to learn, and plenty of after hours discussions to refine what we think we know.
Mike Pallamary, the response I heard to your tip was "I cut it twice, and it's still too short!"
Having been married (note past tense) I found Dave Karoly to be on the money about lunch - I wish I had learned that the cheap way...

William Magee wrote:A proper survey is an accurate survey, which is not necessarily the result of a precise survey.

Now that's a good post.



A couple other tidbits:

It takes less time to do the job right than it does to do it again.


You can learn something useful from everyone. (told to me by one of the first LSs I worked for in answer to a remark I made about some really dumb thing done by a particularly foolish member of the crew)

I retraced a lot survey for a builder I work for to confirm the existing 1" pipes set by another surveyor a couple years before were properly located. He had just bought the lot. I traversed the block, closed and adjusted my field work. After evaluating the data I concurred and agreed with the prior surveyors work and pipes.

About 6 months later I was called back because one of the monuments was missing and a sub built a wall that appeared to be sticking into the road RW. I set up, back sighted and staked out what I shot before. Buzzed with my detector, got a ring and concluded the grading contractor just buried it.. It was flush on the surface when I first located it. I started to dig, no 1" IP, Now I am down 24" and no 1" pipe but still getting a ring. Another 6" and low and behold there is the 2" IP set in 1939 by the original tract surveyor at the BC just where the tract map said one was set in 1939 Never expected to find that.

Morale: when in doubt dig another 6"