Im currently in the process of creating a SOP for structural monitoring to which has lead me deep into equations and discussions with Regional Geodetic Advisers that I've not had since I was back in college. While it has shown that there was in fact some retention had during those years, I've been pondering the question about how to apply these corrections which are not constant.
Due to the way that these structures are setup ( dams ) we're forced to occupy pillars out in the open and locating glass embedded into the structure. This undoubtedly creates a temperature / pressure / humidity difference between the location of the EDM and said target to which is going to effect the measurement result.
Normally, temp is taken at the instrument, ( S9 ) input into the data collector ( TSC3 ) and multiple rounds are run to the back-sight and fore-sights. While the values for curvature and refraction based upon these values are easily calculated by the EDM, over distances of +/- 1400' "down canyon, over rivers, ect.," Im wondering how one would / could go about coming up with a procedure that would allow for "correct applications of correction factors" to be applied resulting in an accurate result?
Any thoughts or examples would be greatly appreciated. But take your time - the view from the ( this office ) isn't too shabby:
I do correct the pressure observed at the standpoint for elevation if the line has a significant difference in elevation. I use the mean elevation to compute the pressure. Other than that, I assume temperature is constant over the site.
Kenneth Robertson of USCE ETL prepared a document in 1979 that proposed the use of ratios. In essence, what he was saying is that lines radiating from a given point are subject to similar atmospheric conditions, so if you measure a reference line every so often, you can use the ratio of the observed distance to the "true" distance and apply that ratio to other observed lines measured to points being monitored.
Here is a pdf of that document, see page number 22 (page 25 of the pdf). Actually, the entire document is a good resource. Of course, EDM's are generally much more accurate today.
Just to add to my previous reply...in the picture you posted, I would set pedestals on either side of the valley (2, 4, ?), and shooting between them would be the reference line. Using Trimble as an example, I would turn rounds using the cross valley stable point as backsight so that each round has a shot to it and all of the other points as well. 4 to 8 rounds using a high accuracy 1"/1 mm gun should give pretty decent results.
mcaanda, post: 412244, member: 3108 wrote: Im currently in the process of creating a SOP for structural monitoring to which has lead me deep into equations and discussions with Regional Geodetic Advisers that I've not had since I was back in college. While it has shown that there was in fact some retention had during those years, I've been pondering the question about how to apply these corrections which are not constant.
Due to the way that these structures are setup ( dams ) we're forced to occupy pillars out in the open and locating glass embedded into the structure. This undoubtedly creates a temperature / pressure / humidity difference between the location of the EDM and said target to which is going to effect the measurement result.
Normally, temp is taken at the instrument, ( S9 ) input into the data collector ( TSC3 ) and multiple rounds are run to the back-sight and fore-sights. While the values for curvature and refraction based upon these values are easily calculated by the EDM, over distances of +/- 1400' "down canyon, over rivers, ect.," Im wondering how one would / could go about coming up with a procedure that would allow for "correct applications of correction factors" to be applied resulting in an accurate result?
Any thoughts or examples would be greatly appreciated. But take your time - the view from the ( this office ) isn't too shabby:
Just an idea: Attach a fast reacting weather meter, like a kestrel, to a drone. Point camera at weather meter and fly the line of sight noting temp along the line vide video screen. Only problem I can think of would be minimum focal dist on a fixed lens camera. You should be able to fashion a lightweight rig to hold the weather meter away from the prop downdraft and hopefully into the useful focal range of the camera
We use a Brunton ADC Pro, it stores the data onboard, and can then be downloaded. No need to have the data in real time, I correct al of my EDM data back in the office.
The drone idea is a good idea...when they were doing crustal monitoring decades ago before GPS they would sometimes fly the line with a fixed wing or helicopter to get weather data along the ray path, needless to say very expensive.
mcaanda
1. What order of Survey ( 1st,2nd,3rd) are you trying to do/write SOP's for ? (this would make a big difference in the rest of our conversation).
A little help can be provider in saying there have been many books/articles written on the subject of Deformation monitoring and Trilateration.
2. What is the max/min distance (horizontal) will you be looking at ?
3. What do you think the max. difference in elevation will be?
If you answer the 3 questions above I will be able to help you better.
JOHN NOLTON
Tombstone, AZ
John Hamilton, post: 412315, member: 640 wrote: Just to add to my previous reply...in the picture you posted, I would set pedestals on either side of the valley (2, 4, ?), and shooting between them would be the reference line. Using Trimble as an example, I would turn rounds using the cross valley stable point as backsight so that each round has a shot to it and all of the other points as well. 4 to 8 rounds using a high accuracy 1"/1 mm gun should give pretty decent results.
There's concrete pedestals that were installed during construction where they started out using triangulation and in the 80's went to trilateration. Flip all kinds of angles to the 4 pillars that are to be occupied, and then shoot out to the glass on the structure. Being that there's going to be a long historical background, the thought to continue to use the trilateration would be the best option as to continue the apples to apples evaluation of movement.
Conrad, post: 412343, member: 6642 wrote: Just an idea: Attach a fast reacting weather meter, like a kestrel, to a drone. Point camera at weather meter and fly the line of sight noting temp along the line vide video screen. Only problem I can think of would be minimum focal dist on a fixed lens camera. You should be able to fashion a lightweight rig to hold the weather meter away from the prop downdraft and hopefully into the useful focal range of the camera
Even w/ the 107 cert, being able to fly any small craft in and near the structure is dang near impossible. I've yet to see an application approved when asked.
John Hamilton, post: 412348, member: 640 wrote: We use a Brunton ADC Pro, it stores the data onboard, and can then be downloaded. No need to have the data in real time, I correct al of my EDM data back in the office.
Agreed - I've never understood why there's so many that I've had to explain why there's no need to have the results right NOW... there's so many variables that are in play that being able to give any sort of definitive answer is playing with fire being that there's so much looking into what you've collected to ensure that there's no issues with the dataset.
And to make matters worse - I always get a call 5-6 months later asking if the info was sent... Um yea - months ago...
JOHN NOLTON, post: 412383, member: 225 wrote: mcaanda
1. What order of Survey ( 1st,2nd,3rd) are you trying to do/write SOP's for ? (this would make a big difference in the rest of our conversation).
A little help can be provider in saying there have been many books/articles written on the subject of Deformation monitoring and Trilateration.
2. What is the max/min distance (horizontal) will you be looking at ?
3. What do you think the max. difference in elevation will be?
If you answer the 3 questions above I will be able to help you better.
JOHN NOLTON
Tombstone, AZ
1/300,000 or better is required per spec.
And yes, there's quite a bit of information out there dealing with deformation and trilats to which I've been going over for quite some time. Actually some pretty interesting information out there which leads ( atleast myself ) down more than a few rabbit holes and many hours seem to have just "disappeared."
Horizontal distances are going to be anywhere from +/- 500' to 2k' max I'd think. I would have to dig into some of the older datasets that were collected to come up with a more accurate value. There "is" thought on another ( not the one specifically in question ) that would have long baselines +/- 5280' that this info would assist with as well.
For the trilateration elevations would be within... say +/- 50' difference? This is a guess, as its different for each structure. But as a "baseline" I'd say that they'd be in that ballpark.
mcaanda
Your 1/300,000 or better........ Am I to infer you mean Closure in position? IF so then you are looking at doing 1st Order Work.
If Yes then you should get a copy of "Classification, Standards of Accuracy, and General Specifications of Geodetic Control Surveys,
by NGS Feb. 1974,reprint 1976. Along with that get Specifications To Support Classification, Standards of Accuracy, and General Specifications
of Geodetic Control Surveys, by NGS, July 1975. Also get Standards and Specifications for Geodetic Control Networks, by NGS, 1984.
You should also get all standards that have been written since 1984 by NGS. Such as standards for doing base lines, GPS, barcode leveling.
Your distances will be short, max. 2,000 feet (609+ meters) but never assume a temperature or pressure over the length of any line by measuring the
the T&P at the instrument only (very bad survey practice). In the NGS publications above (if you do not have them) it clearly states that T&P and
wet and dry bulb readings will be taken at each end of the line.
You will need a very good meteorological equipment. Remember that a 1 deg. C error in air temperature will cause a 1 ppm error. The same is true
for a 3mm error in Barometric pressure; that it will cause a 1ppm error in distance. This is for Laser and Infrared EDMI. The humidity correction
(wet bulb/dry bulb) can cause up to ~ 1.6 ppm. This is usually in the tenths of ppm of a correction but should be measured and corrected for, for the
order of work you are going to do.
Instrument selection. A theodolite(or TS) that has a DIN or ISO (modern terminology) spec. of less than 1 arc second (both Horiz. and Vert)
EDM + or - 1mm or less
JOHN NOLTON
PLS Cecilia Whitaker is the (retired?) expert both optical and later GPS concerning dam deformation and other high resolution studies. She even developed near real time systems that would issue automated warnings to downstream agencies. You might search the interwebs for her papers.
Mike Marks, mcaanda made NO reference to an automated system of monitoring and I have left out all kinds of names because as much as mcaanda is reading I am sure he will find her reference and 20 others that have done papers on automated systems.
One name for sure is Dr. Adam Chrzanowski which is on one of Her papers. All interested can find some information on Cecilia Whitaker in
the California Surveyor , Fall 2008 #156 starting on page 46 (nice read).
JOHN NOLTON
Sorry for the serious delay in my replies - I've been flat on my back counting all the holes, nicks, crooks and crannies in my ceiling in between moments of semi lucid consciousness due to this respiratory infection I've been fighting.
This is 1st order required to which when really looking at all the potential factors that need to be considered in ones error budget, Im hoping that nobody has a sneezing fit during the data collection as it could end up causing a fuss. [/slight sarcasm]
We have recently purchased a new Trimble S9 which spec's out at 1mm +1ppm which I have dictated will ONLY be used for instrumentation. I use an S6 for my daily driver. As for the atmospheric goods needed, I'm still working on that as there's currently nothing on hand which will allow for the needed measurements to be made.
And last but not least, Im putting together a "how to" so that the data collection can be not only standardized but ensure that the results that we're sending out are what we say they are.
Good thing the weekends coming up - I think some rest and reading is in order.
I wish I had a new S9! I am using a high precision S6 (1", 1 mm + 1 ppm) that I bought in 2006, the thing just plain works. I have told the shop that anytime I send it in for cleaning, it is just for that, cleaning and NOT any adjustment to the EDM. I don't trust them to have a baseline accurate enough. We take ours every year to a baseline in Ohio that has pillars rather than disks, so it eliminates centering error.
But if you do an analysis you will probably find that the uncertainty in the distances results in much better point accuracy than the uncertainty in angles. I use both star*net and Geolab, either one can do simulations. You put in the approximate locations, your planned observations, and the accuracy of your observations, and they will give you an estimate of the accuracy achievable. As long as you are realistic with the accuracy estimates, they will give realistic answers. As John Nolton said, you need to be cognizant of the errors due to uncertainty in measuring atmospheric conditions (T, P, RH) as well as centering errors and instrument capability.
As for using the drone to collect atmospheric data, I don't know why it would be prohibited to fly close to the dam as long as no people are below. A drone crashing into face of the dam is going to do ZERO damage to the dam.
mcaanda, glad you are OK now. I have been sick for 7 days with some type of virus and had trouble posting the above post. (1st day out, today).
Remember that all your equipment(when you write up your SOP's) should be traceable to NIST; tapes, level rods, Temp. probe, Humidity probe,
Barometric pressure probe. EDMI to a base line for calibration. Make sure that you keep a log of the adjustment of all other equipment;
tripods, tribrack's, prism's. Number or letter each piece of equipment so its easy to identify and put it in your log book.
Stay Well
JOHN NOLTON
John Hamilton is the Trimble S6 you have a 1 second or 3 second inst.? How much did it cost?
JOHN NOLTON
PS how much does this S9 cost? (could not find it by internet search).
$25,250 at Martin Instrument
The S6 I have was something like $36K back in 2006, plus another $4K or so for the data collector, if I recall correctly. I do not just use it for instrumentation, it is the workhorse for everything we do. I never thought it would last this long. I have wanted a new S9 for a while, but now the SX10 comes out, 1" angles, 1 mm å± 1.5 ppm, so that is now on my want list.
I bought a Trimble M3 (small, lightweight, longer reflectorless range but not robotic) a few years ago but it almost never leaves the office. This picture always makes surveyors cringe...we traversed 25 miles down the middle of a river in 2009. One day we heard thunder off in the distance, but no rain on the river. Then I heard a rumbling sound, and looked upstream to see a wall of water coming down. I put the instrument in its case, jumped in the canoe, and rode the whitewater a couple of miles to where we had left a truck. Flash flood from a storm on a tributary. I thought the dam had opened up its gates, which pissed me off since they knew we were working downstream, but it was not them.
John Hamilton, post: 413362, member: 640 wrote: I wish I had a new S9! I am using a high precision S6 (1", 1 mm + 1 ppm) that I bought in 2006, the thing just plain works. I have told the shop that anytime I send it in for cleaning, it is just for that, cleaning and NOT any adjustment to the EDM. I don't trust them to have a baseline accurate enough. We take ours every year to a baseline in Ohio that has pillars rather than disks, so it eliminates centering error.
Hi John, I'm not sure yours is the best policy, but perhaps it is depending on the capability of the service shop. They may have a micro baseline better able to determine the zero correction than a longer, open air baseline. and they may have a very accurate frequency standard with which to calibrate the frequency generator in your EDM so that scale errors are kept in check. I understand that authorised service shops for the brand I use have these facilities. Perhaps your nearest shop does not?
Who would you trust to adjust your instrument if your baseline measurements were beyond your tolerance?
Nice pic BTW.
JOHN NOLTON, post: 413389, member: 225 wrote: PS how much does this S9 cost? (could not find it by internet search).
We're a federal agency and have to use GSA Advantage for our pricing. Our hands are tied with negotiating so if I remember it was somewhere in the neighborhood of 30K.
Conrad, post: 413560, member: 6642 wrote: Hi John, I'm not sure yours is the best policy, but perhaps it is depending on the capability of the service shop. They may have a micro baseline better able to determine the zero correction than a longer, open air baseline. and they may have a very accurate frequency standard with which to calibrate the frequency generator in your EDM so that scale errors are kept in check. I understand that authorised service shops for the brand I use have these facilities. Perhaps your nearest shop does not?
Who would you trust to adjust your instrument if your baseline measurements were beyond your tolerance?
Nice pic BTW.
Well, they MIGHT have that kind of equipment, I don't know. Years ago they told me they open a window and shoot to a prism across the river, not sure where or how stable, or how they got their "calibrated" distance, so I prefer to err on the side of caution. If it ain't broke...
All of the deformation work I do is processed in the office, loaded into a database, I can apply a scale and offset to the measurements very easily, but have not done so yet on a regular basis, have not seen the need for it. But, a couple of times we had a guy out with us from another company who had another S6. We used both, I used the adjustment software to compute scale and azimuth parameters for his instrument (which was not a high precision unit) relative to ours, and then applied them. They were pretty small, 1 or 2 mm and 1 or 2 ppm if I recall correctly.
I went on a tour of the Trimble (used to be Geodimeter) factory in Stockholm a few years ago. Of course they had lots of laboratory level gear around to calibrate, etc, but they said they also open a window and shoot to various prisms far away, at various distances, just to make sure everything is correct.
John Hamilton, post: 413840, member: 640 wrote: I went on a tour of the Trimble (used to be Geodimeter) factory in Stockholm a few years ago. Of course they had lots of laboratory level gear around to calibrate, etc, but they said they also open a window and shoot to various prisms far away, at various distances, just to make sure everything is correct.
I saw the same thing in Heerbrugg; the permanent targets are across the street in a hillside vineyard.
