I don't think I've ever heard someone call curves railroad or regular.?ÿ heh?ÿ Do you mean spiral or horizontal?
That's what I thought initially, but could also be degree of curvature i.e., 100ft arc for "regular", 100ft chord for railroad...
Perhaps on the next project when I set the original 4 control points I??ll run levels and adjust the vertical on them but when I localize I??ll only hold the first one or one that??s set in concrete for vertical. Not sure how it can get much more precise than that.
There's nothing wrong with doing a vertical shift + geoid calibration (no tilt) and using more than one point to dial in that shift.
I always start with one and check the residuals on the others; if they are more or less evenly spread out I'll add them to the calibration computation and I'm off to the races.
@bstrand now you know why many do not like site calibrations at all. They have a purpose but as many have stated we have moved beyond that today in many ways. One our hybrid geoid models are much better so vertical is not as bad our rtk engines are more reliable than they were many years ago. And the software has made it just as easy to stay on a known datum and coordinate system that calibrating is not the go to answer. ?ÿNow we have counties i am sure all over the country with what they say is nad83 state plane coordinates and the land surveyors have been told they must tie to one of their monuments when doing plat submissions in those counties.?ÿ
i tied to a ngs monument my first time back in private sector off vrs network. I observed it twice once before i hit the project site in the morning and again at end if the day. As i was told to tie to one known monument. So i pulled up ngs found my area read the sheets and printed the ones around my site. Got lucky first one I found hit it nicely hz and vt. Got back next day and was told additional work needed on site and would be going out again at end of the week. For giggles I stopped by another ngs monument on other side of site and same results it was several miles further from site. Did the work. Next week office guy ask why i did not do a site calibration. ?ÿI said no need. He said i must. I said why he said we always site calibrate to county monuments. I said oh well where do I find these county monuments. Remember i came from the book days of looking up monuments on private sector side. Well he sent me a link. I looked around and so oh goody two monuments from ngs I already tied are published here I found more monuments closer and surrounding site and noticed the elevation difference. I said this will not work well explained it to office guy and at same time explained that something was a little crazy with the coordinates published ?ÿby county as they show in report they held those two ngs monuments but the coordinates were way different no knowledge of a scale or combined factor or even a project type factor. It was listed as nad83 2011 etc. i have no idea what they did and told them i didn??t want to know. I said here is all monuments and data proven to meet state standards and tied to ngs values very well.?ÿ
what i did learn later as their practices no joke. In office they would tie monuments then scale to ground in TBC do site calibration export to cad and thats what it was. ?ÿI have read calibration files of several hundred ppms in tilt seen scale factors to make residuals look ok ?ÿseen residuals of points be over half a foot ?ÿ. ?ÿ
Here is one set two rtk points grid ?ÿset robot up on one site other whatever the error.05 ft or .35 ft hold zero re shoot to make distance work between them ?ÿtraverse close back on self ?ÿall at scale 1.00000 not projection and plat as nad83 grid bearing ?ÿbecause we only changed distance not angle so we are good ?ÿi have seen a foot of distance error because trees multi-path but thats ok because we know we are good. Me ask question which one is good how do you determine which of two points are bad. ?ÿI would tell them i would rather see them assume coordinates and forget nad83 and work than do what they did ?ÿheck at least one old timer would scale from map a lat long and note that because he would say i don??t understand gps and don??t need it and i give them what I understand and notes it according. That that lat long was scaled from a 1:xxxx map etc that is more honorable than fudging in my opinion. ?ÿ
?ÿ
An early on calibration job:
?ÿ
Obviously the line between Bench marks isn't perfectly straight but they are moving east, right to left, and not too far from forming a straight line.
"Luckly" for us we live in an area where Geoid heights decrease heading to a mountain at roughly .5' per mile, of course this isn't consistent. ?ÿ
When GPS came out it was recommended to calibrate to facilitate GPS as a valid elevation instrument.?ÿ
What we found out quickly was that the calibration failed no matter how it was applied to a large enough site if you were looking for any type of accuracy.?ÿ
Didn't seem to matter what level control we used it never worked.?ÿ
There was a seminar given by a Trimble rep explaining how to do static and adjustment of a static network. Basically, the idea was to figure out the horizontal then after it's all finished calibrate the network. That was thrown in as the very last task in a two day seminar and all we could ever come up with was that it's a way to make the elevations work. But, by then we all knew they weren't going to work very well, not in the Rocky Mountain region. So, you had horizontal residuals that were perfect and residuals only for the vertical.?ÿ?ÿ
Eventually Geoid03 was released and changed everything. It worked much better than any calibration when applied to a site. Over large areas there were issues but later Geoid Models have continued to refine those problems.?ÿ
The above graph is for the local airport, the errors we were consistently seeing there were in the 0.1' to 0.2' between calibration points, not terrible but unacceptable for runway elevation control. By then we had a dense network of bench marks running all over the site.?ÿ Smaller sites also were analyzed and found to have similar issues, it was very concerning. 130 acres and there was no way to get good verticals out of a calibration, tight elevations surrounding the parcel and running down each roadway inside the 130 acres, but no combination of calibrated points gave good results. There were some attempts to merge geoid models with a vertical calibration, I tried them years ago and they didn't work, haven't tried that with newer models since I gave up on calibrations about 1998 or so.?ÿ
@on_point I guess i am trying to figure out if its a new project and you are establishing control why would you need to calibrate. Calibration or localization is for tying gps to existing control. Surely surveyors are not setting rtk control and then calibrating to that control. If a new site no existing control then just set it use good procedures and redundant measurements at a different time if using base and rover move base and observe others perform least squares. If vertical needs to be tightened up then you can run levels to make all control relative in elevation better. But no need to calibrate as you are not going to improve the rtk measurements by doing so.
@mightymoe yep. I believe at that time vertical rtk precision were plus or minus about.15 ft for rtk forget geoid model just in repeating the vertical. ?ÿWhat part of Rocky Mountains airport. I was in Colorado in 98 as well. The year my first born arrived December 23. ?ÿHe is flying into town today with my oldest daughter. ?ÿWe will celebrate our birthday tomorrow lol. Yes we have same birthday. ?ÿBrought that little sucker home on Christmas day and they had him stuffed in a Christmas stocking.
@mightymoe yep. I believe at that time vertical rtk precision were plus or minus about.15 ft for rtk forget geoid model just in repeating the vertical. ?ÿWhat part of Rocky Mountains airport. I was in Colorado in 98 as well. The year my first born arrived December 23. ?ÿHe is flying into town today with my oldest daughter. ?ÿWe will celebrate our birthday tomorrow lol. Yes we have same birthday. ?ÿBrought that little sucker home on Christmas day and they had him stuffed in a Christmas stocking.
I'm not sure what you mean about 15 ft., the ellipsoid is about a 45 foot smaller number than NAVD88 near the airport, less on the west side, more on the east side. Geoid 90 wasn't a great resource, calibration was your only hope to get close and it was a fail, still would be. So, all elevations were levels or total station derived. I'm assuming people that get it to work do it in areas where the geoid is less messy.?ÿ
I'm heading out tomorrow morning to take my grandson away for a while from his awful parents, they keep telling him all the things he can't do. Don't know why they are so mean. Poor mistreated little guy.?ÿ
?ÿ
@mightymoe 0.15 not 15. Sorry. I mean we could not always repeat precisely of much better with rtk depending on baseline from base and satellite geometry and vdop. ?ÿNow we are much better and the geoid model is around 2 to 3 cm where i am. Not the geoid. ?ÿThe GEOID vs the hybrid geoid model we use for orthometric heights.?ÿ
Yes telling kids what they can??t do is not always the best ????. ?ÿI have tried with my second batch of young ones to not use cannot as much. I might say that is not a wise choice or i would not try that because this might happen etc etc.?ÿ
Those parents need to be more accommodating. He wants ice cream, what's wrong with that, get him the toy for heaven's sake, he wants to be at the park all afternoon, you guys can put off all the chores you think are important.?ÿ
@mightymoe my parents tell me they get to spoil my kids and send them home to me to fix lol. 2 of mine are out and on there own now. Two more to go. Then i get to spoil grandkids lol. I am actually looking forward to that day if i live that long. My mom was tough on us. Not her grand children. She will fuss at me if i correct them to much lol. I say mom you would not let me get away with that. She always sais well you were different. Enjoy them. Have a wonderful visit and Christmas.
@mightymoe my parents tell me they get to spoil my kids and send them home to me to fix lol. 2 of mine are out and on there own now. Two more to go. Then i get to spoil grandkids lol. I am actually looking forward to that day if i live that long. My mom was tough on us. Not her grand children. She will fuss at me if i correct them to much lol. I say mom you would not let me get away with that. She always sais well you were different. Enjoy them. Have a wonderful visit and Christmas.
There's parenting and there's grandparenting. Jump in front of him when he's booking it for the ocean, or the road, that's about it. Otherwise, it's doing what he wants to do.?ÿ
?ÿ
@on_point I guess i am trying to figure out if its a new project and you are establishing control why would you need to calibrate. Calibration or localization is for tying gps to existing control. Surely surveyors are not setting rtk control and then calibrating to that control. If a new site no existing control then just set it use good procedures and redundant measurements at a different time if using base and rover move base and observe others perform least squares. If vertical needs to be tightened up then you can run levels to make all control relative in elevation better. But no need to calibrate as you are not going to improve the rtk measurements by doing so.
?ÿ
I set control with the RTS with no projection at a 1 scale then I localize the gps to the control so they??re both on ??project coordinates?. I was just thinking of taking it a step further and running levels and adjusting the vertical before the localization or just holding one control point for vertical.?ÿ
?ÿ
@on_point Maybe i am missing the acronym RTS. Is that robot total station. If so I got what you are doing. ?ÿWhy not set your control with RTK first and then set up the robot on datum and trig level the elevation as a check and if you see outliers and prove the robot is better relative. Then adjust those points. ?ÿNow a level will always give better than rtk or robot difference in heights. But if you start with gps no need to calibrate. I don??t know what software you run but I imagine most can handle terrestrial data on a datum and projection now days. ?ÿThen your project is on a known datum you can always update at anytime in the future. ?ÿEven in the worst case scenario and you lose every control point on the job your still on a know. Datum and you can come back in and be within the position tolerances and any transformation tolerances in the future. 5 10 15 20 years from now. ?ÿThats the beauty of working on a known datum. ?ÿI can go backwards and forwards with some confidence.
1) Yes RTS means robotic total station
2) I have to calibrate the GPS to match ??project coordinates? so the GPS measurement matches EDM measurements.?ÿ
3) If I were doing boundary then yeah SPCS all they way but contractors like to see the layout match what their tape says.?ÿ
The GPS is having trouble convincing the RTS that the earth is indeed not flat. ?????ÿ
@on_point That is now understandable. ?ÿ How larger are your sites usually. You could do like some here and from what I have read it is the better approach. ?ÿCreate a LDP. You could do this before going to job site. The. You would still have a known datum and the projection you chose and coordinates system all known. ?ÿSet RTK control first. This can be done to make that gps measurements match contractor tape. ?ÿOr and it is a solution. Perform a grid ground scale just if you do please please please make the coordinates false northing and easting off or add some offset to them. Then save those settings and document them . So anyone following you can see what you did. ?ÿBy saving the parameters it give me or anyone else the ability to know how where and such you scaled and from where. Now you on surface based on one or multiple points. The one point thing works for smaller sites and not a lot of relief.
Perform a grid ground scale just if you do please please please make the coordinates false northing and easting off or add some offset to them
I agree, using coordinates that look like SPCS but are not can easily confuse someone so I try to avoid that.?ÿ
Most sites are ~25-50 acres up to ~400 acres so far. Buildings range anywhere up to ~3000????so far. Haha
Most sites are ~25-50 acres up to ~400 acres so far. Buildings range anywhere up to ~3000????so far. Haha
Good sizes for LDPs.
@on_point I guess i am trying to figure out if its a new project and you are establishing control why would you need to calibrate. Calibration or localization is for tying gps to existing control. Surely surveyors are not setting rtk control and then calibrating to that control. If a new site no existing control then just set it use good procedures and redundant measurements at a different time if using base and rover move base and observe others perform least squares. If vertical needs to be tightened up then you can run levels to make all control relative in elevation better. But no need to calibrate as you are not going to improve the rtk measurements by doing so.
?ÿ
I set control with the RTS with no projection at a 1 scale then I localize the gps to the control so they??re both on ??project coordinates?. I was just thinking of taking it a step further and running levels and adjusting the vertical before the localization or just holding one control point for vertical.
Unless there's some significant vertical relief across the site are you sure that leveling would improve much on the robot measurements??ÿ Are the areas you work in fairly hilly?
?ÿ
I??ve used trig leveling on heavy highway projects for over 30 years. Hasn??t been a problem yet. A 1 second RTS works well unless you??re a flat earth-er that insists on measure ups to establish the HI. In hilly terrain it??s probably better than the multiple turning points that are often necessary when using differential leveling.
