Glad to see his name keep popping up here recently.
I knew when I met Earl that he was a special character. Ive learned tons from him and really wish I could have been one of his students at NMSU when he was a professor in the survey engineering program. Indirectly I am, and still learn
Well, I don't know if anybody's mind was changed, or if anything was really resolved, BUT I do know, that this was one of the most entertaining threads in recent months. Kudos to all who participated!
@loyal I came across an old thread this morning that was about grid and ground issues. I was trying to pull some information as i try and do once a week to give to my crew chief for study material as he prepares for the FS exam. I always learn better myself by listening to different instructors with different examples and such so I always teach the same way. One of your statements was about the best and most clear statement i have ever heard. It was simple yet as i am sure you know many get it wrong daily. The CF on OPUS and data sheets are for ground to grid . I am paraphrasing. But i had a tech read it and a manager and the glossy glaze that went over there face said it all. So many because we now start out on grid assume its the grid to ground. Wish i had your way with words for sure. I still believe computing everything long hand makes a difference in understanding. I just can’t get it out in a way for people to understand as they have this whole lets just click in this area and scale from that or 0 0. I say when does that not work how do you know if its just a production process and no checks are made. So i am one who test it around the site from the data to see first. Vs just clicking and such.
Thanks OleMan, I guess I'm like the proverbial blind squirrel that finds a nut every once in a while. A lot of these issues that we discuss are like those things that any 10 year old can do with 20 years of practice.
@loyal muscle memory. Well you are more than a blind nut finder I am sure. It has been a learning curve coming back to working on the surface when I spent so much time on the ellipsoid and such. Now my dyslexia has to be more dyslexic to go the other way again lol One of these days soon We need a LDP 101. I am studying this as much as possible now. I have established projections and new datums in area’s but I always placed everything on the wgs84 ellipsoid. For a reason or two. Then I started positioning satellites and computing URE now i am back to earth and reality trying to soak up everything i can as fast as I can before i kick the bucket. Or my wife kicks it for me. That is more likely as I am stubborn and well just grumpy.
A lot of these issues that we discuss are like those things that any 10 year old can do with 20 years of practice.
I resemble that remark however it took me 30. I kind of lost interest when the discussion turned to NAD27. I suppose some will feel the same way about NAD83 in 2083.
Plane coordinates may not be used at all in 2083, but using only lat/lon/height or XYZ coordinates poses other problems. For example, XYZ pairs produce slope distances and lat/lon/height distances start with ellipsoidal distances which are "adjusted" to ground.
To me, the greatest takeaways from this discussion were:
1. CAD demands coordinates and is a key driver in ubiquitous reliance on coordinate systems.
2. Trimble calculates ground distances from ellipsoidal distances and ellipsoidal heights. The nuts and bolts of how that produces an arc distance rather than a straight-line distance at a mean height remains an issue. It can be resolved empirically by comparing Trimble ground distances to the same distances derived by distances computed by other means.
There's always something else to learn.
17 pages, so far
@mathteacher about "plane coordinates may not be used at all in 2083, but using only lat/lon/height or XYZ coordinates poses other problems." One thing we are seeing is that surveys are no longer freestanding collections of information, with a vague connection to the rest of the world through a location map placed somewhere on the drawing. They are tied into property tax systems, E-911 maps, culvert maps, wetland maps, etc. But the vast majority of lots have not been surveyed in the last 50 years. So how well will XYZ surveys play with older adjoining lots that were surveyed with plane methods? Or, just approximated by a tax map drafter?
That's been an ongoing issue. I was retracing an older set of ROW's circ 1930's. They didn't bother to set ROW monuments and I was having lots of trouble making sense out of the geometry. It's a US Highway running NW-SE. I just couldn't get things to come together,,,,,,,until it dawned on me. That highway is all true north. As I worked SE from the beginning I simply had to adjust the bearings (usually through the curves) and it fit together very well.
Same with RR right of ways. So Lat, Long, true bearings, surface distances were the norms on many surveys prior to coordinate systems. The old guys had little problems with it, it's the newer ones like me that did. It's illustrative when I got calculation sheet from a 1930's era survey. The surveyor was working for a private energy company but was GLO trained. All the numbers were based on northings and westings and were clearly Lats, Longs probably using a Solar Compass or more sophisticated instruments for the time period. Very complicated and very intense calculations. Eye opening, I don't think shifting to real geodetic will be that big of a deal if we get industry wide computer programs supporting it.
@mightymoe "I don't think shifting to real geodetic will be that big of a deal if we get industry wide computer programs supporting it." The trick is to place a broad enough meaning on "industry wide". ESRI would have to be in on it at a minimum. I wonder how many customized tax map packages are out there.
Surveyors, unfortunately, tend to have a "one-and-done" mentality when it comes to their work. Part of it is because for most of our licensed work, the subject boundary only needs to relate to itself, so if it took extra time and effort to relate it to a global datum, we weren't interested.
Part of it is also a little self-serving and protectionist - if we do tie our work to a global datum and retain that information for ourselves, we have a jump on competitors down the line, as well as any GIS "cowboys" (not all of them, but certainly some) who would want to treat a shapefile of our linework as gospel.
The other problem is QA/QC - there are many ways to tie to a global datum, and many ways to report (and fudge or screw up or misinterpret) results. Do we lock down procedures (think NGSIDB) or take a hands-off approach?
@rover83 When it comes to GIS folks (cowboys or otherwise) treating a surveyor's shapefile as gospel, it seems somewhat justified. As far as I can tell, in Vermont, the only general coverage publicly available database of property shapes is the tax map parcel coverage. As far as I can tell, the first choice for a shape is a survey map and the second choice is a deed description. Lacking those, some boundary is chosen that prevents overlaps and gaps. I don't know what the process is for placing and orienting these shapes. Aerial photos or the E911 layers containing GPS locations of road center lines seem like good guesses.
As I see it, the downside is that to put a shapefile into some systems such as ESRI, the provider has to relate it to a global datum, or it doesn't go in at all. The quality of ties to the global datum will depend on what field procedures were used and whether the details survive. It would be a shame if GIS folks treated all surveyor shape files the same no matter how they were related to a global datum.
@mightymoe makes a good point about calculations in the old days and you make a good one about somehow fitting standalone surveys into a world geodetic system. It's hard to see how the latter can be accomplished without a sound understanding of the former. Whatever understanding there is seems to have been relegated to software.
I do the same thing with my income taxes. I have no idea what my marginal tax rate is, I just assume that TurboTax knows.
The big problem that I see with 3d coordinates is that there's likely always to be demand for flat maps, and those are 2d. My county GIS images are 2d (acually NC State Plane) as is Google Earth. The NC state rosd map is the NC State Plane at something like 1 to 830,000 scale.
Making a 3d map involves a decision about perspective. Are we viewing it perpendicularly from above, horizontally at 180 degrees, or something in between?
If we use lat/lon/height (xyz in instrument speak) it's easy to use an average height. Then distances are very close to measured ground distances and azimuths are geodetic (I think).
If we use XYZ (the 3d system that GPS uses) height is a component of all 3 coordinates and all calculated diatances are slope distances. Azimuths are geodetic.
I'm neither smart enough nor knowledgeable enough to even consider a transition to either. Perhaps that's one of the reasons that NGS is sticking to State Plane.
