Calibrations

:good:
My only question...

What kind of beer?

So on the calibration poll we can put you down as one no, one against?

The only thing I used it for was calibrating to found points on a record map then using it to stake out and find more points but ultimately the calibration was dumped and the project adjusted within itself.

If I can't place gps and a total station on the same points and get the same results, I start making phone calls to other vendors. My Dad told me a foot is a foot no matter what you measure with.

Davis118 your Dad told you right. No need for redefining the foot. Same foot with GPS as it is with total station.

I much prefer the term "localization". Just makes more sense and is less confusing.

Loyal,
I agree, mostly. HOWEVER, sometimes you have to play to the customer. If your DOT wants a road staked to THEIR control and elevations, then I would hold my nose and "Calibrate".
Dave

davidgstoll, post: 325750, member: 6999 wrote: Loyal,
I agree, mostly. HOWEVER, sometimes you have to play to the customer. If your DOT wants a road staked to THEIR control and elevations, then I would hold my nose and "Calibrate".
Dave

I hear ya Dave, and agree with your statement. My experience with DOT(s) is pretty much limited to UDOT (Utah). I have seen UDOT do things MORE than one way, and some "techniques" are more sophisticated than others. Over the years I think that 'they' have tried to become more 'standardized.'

Loyal

A couple of days ago we surveyed a couple of rural lots in a subdivision we surveyed about 10 years ago. I developed an LDP at the time. None of the boundary monuments we immediately found were suitable for a base station so I set it up autonomously. I tied to one monument and performed a single point localization. Because the orientation was defined by the projection, we were able to navigate to the remaining monuments within a couple of centimeters. Later I performed another localization using all of the points I found. We've been doing 3D on all of our surveyors for several years so I had valid heights on all points. The localization reported that I had a rotation of 8 arc seconds. Scale of 23 ppm. I turned of rotation and scale. This resulted in a translation only localization.

Rule 1 of localization is know your control.
Rule 2 is carefully evaluate parameters - rotation translation and scale.

I generally believe there are better tools than localization but localization when done expertly, can perform QCQA and quickly relate a current survey to a legacy survey.

Bingo. I would rather have a crew localize and know what they are looking at than fake thier way through using a 'superior' method and leave incorrect pins in the ground. ..

Loyal, ONE POINT CALIBRATION, the term itself is plainly an impossibility, I have tried to remember the first time I ever heard such a thing and I think it was here on this board. I had never heard of it before.

As an aside I haven't had to calibrate to a DOT job since 2000, and that particular job was only because DOT insisted on keeping old NAD27 control as they didn't have time to refly it and there was existing mapping, otherwise since the early 1990's all DOT projects have been based on static NAD83 control so, therefore, there is no need to calibrate. Also, when I do see people calibrate to such control they ACTUALLY calibrate using RTK!!!!!! Think of the implications of THAT!!!!!

At least if you are going to calibrate to known control use static for goodness sakes.

"Catch 2022". Yippee.

Loyal, post: 325708, member: 228 wrote: Rant on...

The term “one point Calibration” is a contradiction in terms in the third degree!

rant off
flak vest on
going out for a beer
Loyal

I was hoping you'd chime in. You're probably the resident expert on this.

gschrock, post: 325878, member: 556 wrote: I had an instructor once who said a basis of bearing from a local system or records is a type of localization, and that a closed traverse where you apply published coordinates to held points is a form of localization. Not sure if I completely drink that... but if we look at a localization as first determining the difference between what the GNSS (preferably long and properly handled static) actually observes at that time, and what Squampish County established as their "god points" decades before utilizing lord knows what assumptions and external reference if at all - those that they insist we tie to.

gschrock, post: 325878, member: 556 wrote: There are still places that insist on for example things like NAD27, NAVD88, thier own vertical or horizontal "datums", or worse still their own realizations however flawed that they insist on calling them something they may only be loosely related to. Choices: do a proper localization or turn down the contract.

If you see work that I had a hand in, you will sometimes see a slightly tongue in cheek note: "SPC WA South 1983/91, City of XXX datum" I haven't been brave enough to just call it the "City of XXX GIS datum of 1997".

gschrock, post: 325878, member: 556 wrote: Hi Loyal,

I hope you found that cold beer and enjoyed it.

If you talking about a fugarwe "calibration" then I would agree wholeheartedly and buy you many beers in ritual affirmation.
But I would have quibble a bit to some degree, and if the subject is finding the optimal ground between the sloppy extreme on one end, and the client-mandate-thus-need-to-do-a-proper-registration-to-existing-references on the other - it is matter of degree. For lack of a better term, lets call it a "localization". I am sure we could all come up volumes of good reasons why localization could or should be relegated to extinction, but there are a lot of instances (I did not say "good reasons", only instances) where some kind of localization must be performed (short of refusing to do the work).

If one is fortunate to never ever have to be subject to such instances, then that is wonderful, but sometimes there is no avoiding it. I had an instructor once who said a basis of bearing from a local system or records is a type of localization, and that a closed traverse where you apply published coordinates to held points is a form of localization. Not sure if I completely drink that... but if we look at a localization as first determining the difference between what the GNSS (preferably long and properly handled static) actually observes at that time, and what Squampish County established as their "god points" decades before utilizing lord knows what assumptions and external reference if at all - those that they insist we tie to.

There are still places that insist on for example things like NAD27, NAVD88, thier own vertical or horizontal "datums", or worse still their own realizations however flawed that they insist on calling them something they may only be loosely related to. Choices: do a proper localization or turn down the contract.

So lets say some client says they want things on SPC Some-Zone NAD83/xx (so many things wrong with that kind of statement in a project spec, but there are some that do that, so humor me). Or a project has been set aside for many years since the design survey (fun with funding) and you have to go back now and add more or stake, or whatever. Their control was based on some published realization of SPC, but they've been holding the same values published during the magical HARN era (decades past). In relative terms their control seems to hold together in an unto itself, but you'll be working with ECEF GNSS. The disparity between what you can observe now, on current epoch, and what the flawed reference framework was decades ago must be accounted for.

Before we consider epoch jumping with the various time-dependent tools - those are models - and if the magic HARN from back in the day had a lot of noise to start with (very true in our neck of the woods) plus the velocity models are not very consistent then that is not a good idea. Then one must get a (properly executed) observed current set of values and compare that to the "what-the-fiduciary?" published values. Hey, it ain't pretty, but in such cases folks have to do the best they can.

Here in places we have had about 0.5' of displacement (yes even in NAD83 "plate fixed") since the magic NAD83/91 realization that so many jurisdictions and enterprises still hold, believing or wanting to believe that their holdings and infrastructure are frozen in space and time. Yeesh. As the NGS reference frame change approaches, "Catch 2022" will see velocity and legacy system affecting folks who have thus far been able to avoid the subject in other parts of the country (places like here have always had to deal with such things to some degree).

If only that that 4th "D" is truly and properly understood applied- maybe then we could see world where localization would be moot. But there is a long road ahead.

"Choices: do a proper localization or turn down the contract."

Or occupy the control and use it as it's given to you, which is what the DOT would want to see, use the NAD83/93 as NAD83/93, that way the lat, long's are connected to the given X, Y and there is no need to add a calibration. It will be odd to send in XY coordinates that don't match the Lat, long of a NAD83 projection

As far as elevations go, NAVD88 isn't going anywhere for a long time, it took over a decade to get new FEMA maps and cause of all the controversy associated with them it will take even longer before counties and cities will sign off on a newer version.

And since those maps are NAVD88 new subdivisions, sewer projects, roads, bridges, will no doubt follow along for a long time.

I'm a bit curious why anyone would calibrate a good control network that was GPS static surveyed in NAD83 regardless of the EPOCH.

gschrock, post: 327056, member: 556 wrote: But the earth moves......

[MEDIA=youtube]wbQ4m-NqeF8[/MEDIA]

gschrock, post: 327056, member: 556 wrote: But the earth moves. If the published value is 25 years old and the today value (observed by long static or say OPUS) is a half a foot different due to 25 years of plate velocity, how do you account for that? Maybe there has been no velocity where you are working but we deal with cm per year in some parts of this state.

Yes, there is movement here also, but why try to chase that? Every few years there is new updated coordinates, and that's the way it should be, but control for projects will always be done before the completion of the project, sometimes many years before.

Therefore, the control needs to be used as it is, there is no way to update each time a new EPOCH is released.

So why not use it as is, accept the LLH and XYZ (GEOIDXX) numbers given, no calibration needed.

Trying to use the new EPOCH is like the dog chasing his tail.

Of course you may be discussing network RTK and it seems that may be one of it's disadvantages, there isn't anything like that where I work.

To me it's unfortunate that the topic of calibrations, Localization, One-Step, Two-Step etc. is even an issue at all and demonstrates the extent of the inroads that GNSS manufacturers have made into controlling how surveyors do their work. Unfortunately, many if not most, surveyors who are using GNSS, particularly in RTK mode, have a very limited understanding of the geodetic concepts involved in positioning with GNSS. So they really don't have a grip on transformations of any sort, and don't understand how to fix their project to local control.

It seems to me that if surveyors understand the broad subject of transformations much better than they do that this problem would have never arisen. All the above methods are variations of different types of Helmert transformations that estimate varying numbers of parameters. In some of the methods the vertical is dealt with separately relative to either a plane or preferably to a geoidal model. If one understands the mathematical concepts, it is not difficult to decide which approach to use for different situations. However, surveyors don't understand the concepts, so manufacturers have stepped into the void and come up with their own obfuscating terminologies that sort of "dumb down" the process so surveyors with limited understanding can get on with their work. Unfortunately manufacturers are very cagey about revealing exactly what their processes are doing in much more than a general sense so the end user is left puzzling over terminology within the software and the differences between what the multitude of terms mean in Leica, Trimble, Topcon and anyone else's software.

So it's kind of a mess and currently now as a training/support guy I get some strange calls from folks out there who should not be charging money for the work they are doing because they don't know how to accomplish what they tell their clients they can accomplish. It's easy to blame the manufacturers, but ultimately the surveying professional using the equipment are often not prepared for the work they are doing. Surveying education can and does help, but many working surveyors also have to push themselves a lot harder and learn some conceptual matters that "good ole boys" sometimes don't like to take time to learn.

Fixing a project to some sort of local control is fundamental to most surveying projects and if surveyors cannot accomplish that confidently then they are ultimately the ones who are responsible for fixing the situation. It is their responsibility and it is disingenuous to try to lay the blame elsewhere.

RCliffWilkie, post: 329694, member: 10285 wrote: To me it's unfortunate that the topic of calibrations, Localization, One-Step, Two-Step etc. is even an issue at all and demonstrates the extent of the inroads that GNSS manufacturers have made into controlling how surveyors do their work. Unfortunately, many if not most, surveyors who are using GNSS, particularly in RTK mode, have a very limited understanding of the geodetic concepts involved in positioning with GNSS. So they really don't have a grip on transformations of any sort, and don't understand how to fix their project to local control.

It seems to me that if surveyors understand the broad subject of transformations much better than they do that this problem would have never arisen. All the above methods are variations of different types of Helmert transformations that estimate varying numbers of parameters. In some of the methods the vertical is dealt with separately relative to either a plane or preferably to a geoidal model. If one understands the mathematical concepts, it is not difficult to decide which approach to use for different situations. However, surveyors don't understand the concepts, so manufacturers have stepped into the void and come up with their own obfuscating terminologies that sort of "dumb down" the process so surveyors with limited understanding can get on with their work. Unfortunately manufacturers are very cagey about revealing exactly what their processes are doing in much more than a general sense so the end user is left puzzling over terminology within the software and the differences between what the multitude of terms mean in Leica, Trimble, Topcon and anyone else's software.

So it's kind of a mess and currently now as a training/support guy I get some strange calls from folks out there who should not be charging money for the work they are doing because they don't know how to accomplish what they tell their clients they can accomplish. It's easy to blame the manufacturers, but ultimately the surveying professional using the equipment are often not prepared for the work they are doing. Surveying education can and does help, but many working surveyors also have to push themselves a lot harder and learn some conceptual matters that "good ole boys" sometimes don't like to take time to learn.

Fixing a project to some sort of local control is fundamental to most surveying projects and if surveyors cannot accomplish that confidently then they are ultimately the ones who are responsible for fixing the situation. It is their responsibility and it is disingenuous to try to lay the blame elsewhere.

Daaaaaannggg. +1 Cliff

RCliffWilkie, post: 329694, member: 10285 wrote: To me it's unfortunate that the topic of calibrations, Localization, One-Step, Two-Step etc. is even an issue at all and demonstrates the extent of the inroads that GNSS manufacturers have made into controlling how surveyors do their work. Unfortunately, many if not most, surveyors who are using GNSS, particularly in RTK mode, have a very limited understanding of the geodetic concepts involved in positioning with GNSS. So they really don't have a grip on transformations of any sort, and don't understand how to fix their project to local control.

It seems to me that if surveyors understand the broad subject of transformations much better than they do that this problem would have never arisen. All the above methods are variations of different types of Helmert transformations that estimate varying numbers of parameters. In some of the methods the vertical is dealt with separately relative to either a plane or preferably to a geoidal model. If one understands the mathematical concepts, it is not difficult to decide which approach to use for different situations. However, surveyors don't understand the concepts, so manufacturers have stepped into the void and come up with their own obfuscating terminologies that sort of "dumb down" the process so surveyors with limited understanding can get on with their work. Unfortunately manufacturers are very cagey about revealing exactly what their processes are doing in much more than a general sense so the end user is left puzzling over terminology within the software and the differences between what the multitude of terms mean in Leica, Trimble, Topcon and anyone else's software.

So it's kind of a mess and currently now as a training/support guy I get some strange calls from folks out there who should not be charging money for the work they are doing because they don't know how to accomplish what they tell their clients they can accomplish. It's easy to blame the manufacturers, but ultimately the surveying professional using the equipment are often not prepared for the work they are doing. Surveying education can and does help, but many working surveyors also have to push themselves a lot harder and learn some conceptual matters that "good ole boys" sometimes don't like to take time to learn.

Fixing a project to some sort of local control is fundamental to most surveying projects and if surveyors cannot accomplish that confidently then they are ultimately the ones who are responsible for fixing the situation. It is their responsibility and it is disingenuous to try to lay the blame elsewhere.

"So they really don't have a grip on transformations of any sort, and don't understand how to fix their project to local control. "

Very often these days we are jumping onto GPS control, it may be local coordinates or some well defined projection or even a "modified" projection such as expanding a state coordinate system to surface, but there are projection parameters associated with the xy coordinates already, what is unfortunate is that the next surveyor then comes along and "calibrates" to it, thus layering another set of tweaks and changes to an already defined system. There are actually some that use the calibration process to "check" each morning, so each day another layer is placed over a good network.

Just like boundary surveys that follow the footsteps, control retractments should accept the parameters already established, until a "real" problem is found to reject it. And that real problem does not include an inability to survey to older GPS systems because the equipment is lacking.