RFC also note the distances Landbutcher shows on his map are grid distances or State Plane distances. The distances you show and that your surveyor showed are ground distances and that is why they look different but are actually the same. Those distances are just being expressed in different datums.
Not sure I understand "they look different but are actually the same". If I understand correctly, I can bring the coordinates back into my survey, do COGO to my heart's content and not worry about making the grid to ground adjustments. Or do I?
Also, I've confirmed that all of my survey and subsequent Star*net work was using US Survey Feet, so I think we can put that to rest as a source of any discrepancy between the two surveys.
When doing an RTK survey you need 2 antennas if you use your own base. Likely the base antenna height was .62' and the antenna he had on his rod was at 2 meters which is the 6.562' height.
I'm guessing he measured from the top of your steel pier to the bottom of his antenna or he may have measured to a specific measuring mark on the side of his antenna
I may be getting in over my head here (nothing new there). So what would I put in the OPUS field for "antenna height"?
Of course, the real problem is likely that the two surveys indeed did not start from the same base. That's indicated by the minimal rotation and the near-constant northerly offset.
If I were RFC, I'd check to be sure that 100 and 500 are the same point.
Of course, the real problem is likely that the two surveys indeed did not start from the same base. That's indicated by the minimal rotation and the near-constant northerly offset.
If I were RFC, I'd check to be sure that 100 and 500 are the same point.
I think a reset may be in order. I've read back through the thread and it seems like I've mistakenly led my friends here to believe I'm trying to account for the 9' and change in location between the surveys. NOT THE CASE! I've never believed that my coordinates were correct. They were a guess! I probably started with Lat/Long from Google Earth, then used NCAT to transform to Vermont Grid zone 4400. The nearest legitimate control point was over 2 miles away and I never wanted to traverse along a busy highway to our property solo. Didn't have my "Prism Girl" to help me at the time.
So, bottom line: my GPS surveyor's point 100 is absolutely the same point on the ground as my 500. It's a steel pier I put in before I started my survey, anchor bolted to solid ledge and has never moved a hundredth let alone 10'.
Now that I have the two geolocated properly, my intention is to analyze mine in comparison to a "real" survey. Now that I have the .rw5, I may be able to run it through OPUS if I can figure out how to do that, then add it to my extensive Star*net adjustment file. Lot's still to learn.
All that said, I am somewhat optimistic that the "minimal rotation" (20") stems from my gazillion celestial shots.
No, your description is great, I just sometimes skim rather than read.
Dropping a pin on GE and reading lat/lon is subject to two main error sources. The significant one is precisely dropping the pin. The second is related to Web Mercator, the ellipsoid GE uses to display it's imagery.
Web Mercator is a sphere with radius equal to the semi-major axis of the WGS84 ellipsoid. GE converts its display coordinates to WGS84 coordinates to within sub-meter accuracy before it's displayed. However, there is a built-in north bias in the Web Mercator display coordinates along with a much smaller east bias.
The 9-foot north and 3-foot East differences fit nicely into that source for differences.
I'd say that your prior work was very, very good. It also underscores what self study along with help from some very generous pros can accomplish.
So remember that 0.3' possible error in that 239 foot line? I calc'd a new #1:2 point 0.28' on the same NW bearing from #2:2 = #5:2. Then I inversed a new NE bearing from #4:1 to #5:2. Check the new map. IMHO you no longer need to rotate your data. This map and points are all the same as your original point list except for the new point #5:2. I did not adjust the elevations because there is too much error. I did not translate anything either until you look over the bearings and coord differences and decide what you want to hold.
I'm trying to wrap my head around what you did and what it means.
Are you saying that if I translate but not rotate my survey, then my original point 3102 (as well as the other 3 corners are in the correct location and should be held, but GPS surveyor's new point 5:2 is off by .28' and should not be held?
As for elevations, I'll deal with that later. Separate subject lol.
I'm trying to wrap my head around what you did and what it means.
Are you saying that if I translate but not rotate my survey, then my original point 3102 (as well as the other 3 corners are in the correct location and should be held, but GPS surveyor's new point 5:2 is off by .28' and should not be held?
As for elevations, I'll deal with that later. Separate subject lol.
Yes I think the old surveyors point 1:2 is off by 0.28'. I am assuming that because the new point 5:2 will inverse to the surveyors old 4:1 and get the exact same bearing as your traverse.
I gave you a snippy reply and I want to apologize for that. No excuses, just an I'm sorry for that.
As you've pointed out, the slight rotation is insignificant, so I did something simple. I regressed the surveyor's coordinates on RFC's, doing Eastings and Northings separate.
The best fit for Northings is just a multiplier with no constant term, like a scale factor. The Easting has a small constant term. It seems that the Easting exhibits rotation while the Northing doesn't, which doesn't make a lot of sense.
In any event, the spreadsheet is attached.
Yes I think the old surveyors point 1:2 is off by 0.28'. I am assuming that because the new point 5:2 will inverse to the surveyors old 4:1 and get the exact same bearing as your traverse.
For clarity's sake, "the old surveyor" is neither an old surveyor nor is the survey old. It's fresh off the press last week. He's in his fourtys.
I am however, really perplexed at how state of the art Leica RTK equipment can produce a point that's .28' off in fairly benign conditions. OPUS won't likely fix this I assume.
One thing's for sure though: I have at least a couple weeks of homework to do to absorb all of the information so kindly provided in response to my query. I really appreciate it guys. I've waited a long time for this moment (of having a legitimate survey) to evaluate my work.
Not sure I understand "they look different but are actually the same". If I understand correctly, I can bring the coordinates back into my survey, do COGO to my heart's content and not worry about making the grid to ground adjustments. Or do I?
Vermont uses a plane that is below the elevation of the ground were your property is. Landbutcher and I gave you coordinates that exist on this plane. When you draw a line from the center of the earth to your corner on the ground, this line intersects the State Plane prior to connecting to your point on the ground. It is the same state plane coordinate at ground level and where it intersects the state plane. You can see in my rudimentary sketch that the distance between 2 points is longer on the ground than on the state plane. Those distances on the screenshot you originally posted are ground distances. The same distances that you will measure. Land butcher gave you the state plane distances between your points. That is why his distances are shorter than what you and the surveyor have measured on the ground.
@rfc I believe you can do simple Helmert around the centroid no scaling on either side. This would show the best fit of your points to his points and maintain both systems creed ability. I think StarNet can do this without fixing a point itself as the translation point and not holding any fixed bearing or line. This will give you the big picture. As far as the 20”. Let’s see probably GPS rtk work. So let’s say he had an error ellipse on two points of .03 ft at each end that depending on distance could account for that if they were in opposite directions. But with rtk gps he has an error ellipse at every point and is usually random error is a given. Did the surveyor locate the same points new than once. So he or she has error in their measurements. Then you have error as well in yours.
I may be getting in over my head here (nothing new there). So what would I put in the OPUS field for "antenna height"?
If you are submitting static data collected by the antenna sitting on 100/500 then it is likely an antenna height of 0.62 will produce the elevation of that point when you convert the RW5 to Rinex and submit it to OPUS. If you are submitting static data collected from the antenna that the surveyor was carrying when he went to measure all of your corners, then inputting an antenna height of 2 meters or 6.562' will likely produce the elevation of the point on the ground. My guess is your RW5 file does not contain any static data but only the data stored in the data collector while he was surveying in RTK mode. Static data is usually collected and stored in the antenna for downloading and processing later. Though sometimes the equipment will be set up to store this data in the data collector and perhaps you do have it in the RW5 file.
If I understand correctly, I can bring the coordinates back into my survey, do COGO to my heart's content and not worry about making the grid to ground adjustments. Or do I?
If you are going to do cogo you need to inverse at ground or at state plane. The distances you measured on the ground are longer than the distances represented by state plane coordinates. Doing cogo and mixing state plane information with ground information will produce erroneous results.
I gave you a snippy reply and I want to apologize for that. No excuses, just an I'm sorry for that.
As you've pointed out, the slight rotation is insignificant, so I did something simple. I regressed the surveyor's coordinates on RFC's, doing Eastings and Northings separate.
The best fit for Northings is just a multiplier with no constant term, like a scale factor. The Easting has a small constant term. It seems that the Easting exhibits rotation while the Northing doesn't, which doesn't make a lot of sense.
In any event, the spreadsheet is attached.
Nice excel file and I could follow everything except those constants a1,b1,a2,b2. How did you get those?
And no worries about your previous reply. I took no offence but thanks anyway.
@rfc I am however, really perplexed at how state of the art Leica RTK equipment can produce a point that's .28' off in fairly benign conditions.
Occasional 0.2'-0.3' jumps in GPS data are one of the hazards. I think it has to do with losing or gaining a satellite or electrical interference and not the equipment. To guard against those jumps I re-shoot my GPS control the next day and later when I start tying in existing property corners with my robot that GPS control gets checked again. Out of 20-30 points I usually find 1 or 2 points that are out 0.2'-0.3' and it is always E-W never N-S.
Below is a screenshot from Geogebra, a math tool. It shows RFC's Eastings as the x-coordinate and Surveyor's Eastings as the y-coordinate. NOTE: There is no surveying interpretation for this. I did it to see the relationship between the two Eastings and then the two Northings.
If RFC's Eastings and Surveyor's Eastings were identical, the points would all fall on the line y = x. They appear to do so, but that's just scale interfering with visual. They are very close, though.
What I want to do is develop a linear equation of the form y = a + bx so that I can calculate Surveyor's Eastings from RFC's Eastings. Geogebra will do that, but I did it in Excel.
Below is a screenshot of the Excel spreadsheet. Note in the ribbon that we are under the Data tab. Over to the right of the ribbon are Data Analysis and Solver. Solver is the add-in that does the least squares line fit that I want.
Below is a screenshot of Solver with entries to do the calculation of the two constants, a and b, that create the equation of the line that I want.
The selected cells are changed iteratively until the criterion, sum of squares of differences = Minimum. It happens in a fraction of a second.
Now I have a linear formula for converting RFC's Eastings to Surveyor's Eastings. Following the same process for Northings gave me a similar formula for Northings.
This works well here because both surveys are on State Plane. We get no statistics on rotation, translation or anything else; that's all bundled into the formula's constants.
The formulas come from a sample of 5 points, so whether inconsistencies in other points is unknown and carries potential to make the formulas inoperative over all points. But that's true of a Helmert transformation based on a sample of points as well.
Crude, based on sound mathematics, perhaps usable, but not strictly professional surveying work. It was fun to do, though, so that should count for something.
So remember that 0.3' possible error in that 239 foot line? I calc'd a new #1:2 point 0.28' on the same NW bearing from #2:2 = #5:2. Then I inversed a new NE bearing from #4:1 to #5:2. Check the new map. IMHO you no longer need to rotate your data. This map and points are all the same as your original point list except for the new point #5:2. I did not adjust the elevations because there is too much error. I did not translate anything either until you look over the bearings and coord differences and decide what you want to hold.
If I understand you correctly, I just need to translate my coordinates (including 5:2) to my surveyors coordinates at 100/500, but not rotate the survey; Then bring those new coordinates into my survey. Is that correct?
It was fun to do, though, so that should count for something.
Isn't that the truth and thanks for the explanation. Your math skills are way beyond mine but I too find it fun to work out a math problem. It's probably some kind of sickness 🙄.
In your latest Excel file, if you were to decide that the surveyor's point 1:2 was an outlier, can it be left out of the calculations you did and still adjust all 5 of RFC's points based on just 4 surveyor's points?
If I understand you correctly, I just need to translate my coordinates (including 5:2) to my surveyors coordinates at 100/500, but not rotate the survey; Then bring those new coordinates into my survey. Is that correct?
Almost correct. Leave out 5:2. That is the surveyor's point 1:2 that I adjusted 0.28'. After translating your 5 points based on moving 500 to 100, you will notice that your new point 3102 will be very close to 5:2 and 500 and 100 should be exactly the same.