California Surveyors Forums

If a point is occupied with gps during an RTK survey, and the survey is calibrated (localized) using one point only, Does that make the bearings observed for that survey "geodetic" bearings? or at least very cose approximation ?

I SAY the "GPS" bearings measured this way are are very close to geodetic,
The other side says they are random bearings and don't relate to anything until a second known point is observed and the survey is calibrated (localized) to two points.

12-pack of Diet Dr. Pepper is at stake here.

By holding only one point the azimuth determination would be based upon the control used, in the case of GPS, from the satellite orbits that you are using, most likely the broadcast orbits which are based upon a realization of WGS84.

It may be close enough depending on what kind of work that you do. I would not call it random because it is referenced to a known datum.

With RTK you have a base and a rover and therefor you have 2 points on the ground. The question is, what is your base referenced to? Are you doing a here survey, autonomus position, processing the base againast opus or is this a one point calibration dealeo goin on and you have assumed 5000,10000, 100 for the base? Whether or not you are on a geodedtic basis depends upon how you have set up the survey. You could be on a local system. Steve's answer was based upon his experience of how a sane person would do RTK. If for example you used a base provided by a commercial service your values could be local values from a calibration or state plane coordinates and the inverse from the base to rover would be a grid bearing....

need more info about proceedures to say for sure which is what....

I asked this very question to Michael McGee years ago. I consider him to be somewhat of a GPS expert.

His response was that the differences between the two are so small, that it is negligible. The difference being a few 1/10ths of a second (The laplace correction, I think).

The subject came up because I was stating true north based on the GPS satellite constellation for my basis of bearings on an R/S and the county was not convinced. Now I use it all the time, especially for public lands work.

My basis of bearings reads something like this:
"The basis of bearings for this survey was true north based on the GPS satillite constellation taken at the South 1/4 corner of Section 34, etc."

Be sure to state where the base station was set since, at my latitude, true north changes about a minute a mile in the east-west.

Mike Stanton, PLS 5702
San Luis Obipso, CA

If you using RTK with one point (using "here" command and not tying out another known point) your bearings are in CCS83 (non-geodetic) based on the satellite constellation (the ephemeris is encoded in the signals but technically can be further refined later by applying a precise ephemeris in an adjustment). You can obtain a geodetic azimuth by applying the mapping angle at that Lat and Long.


One statement that was a little confusing: "I SAY the "GPS" bearings measured this way are very close to geodetic".

Actually these bearings are in SPC not geodetic. However, applying the mapping angle at that place will then result in a geodetic azimuth.

The logic behind this is as follows: You’re at a station and your back sight is the satellite constellations; your software is able to process the signals and do a least squares adjustment for the bearings on the fly. If you were to ever check your work using CORS (a good practice since it’s free information) you would find them to be within seconds of the RTK only solution.

Maybe I'm confusing Geodetic with "true" north.

I was thinking the "here" method (regardless of local coordinates), would provide a very close approximation to "true north" bearings.

For a small survey of less than half a mile, the variation of the north direaction would be insignificant. ???

It sounds like my buddy and i were both wrong.

The Dr. Pepper will be luke warm.

I think the other side owes Good a COLD 12 pack of Diet Dr. Pepper.

Although not necessarily related to Good's predicament, here's my two cents on some of the terminology that's being used here:

Geodetic north equals true north. The Laplace corection (deflection from the vertical) must be applied to astronomic north (solar,stellar determined) in order to get geodetic north bearings. Depending on your location the Laplace corection can be significant for precise geodetic work. I've seen values of over 15" in mountainous areas, as published on the NGS data sheets.

According to Trimble: “If you do a 'here' position in a no projection no datum job you will essentially have azimuths related to true north. As soon as you do a calibration with two or more points or switch to a State Plane system the azimuths will relate to the grid north.”

It takes the mapping angle at the position where you did the “Here” command and applied that to the coordinate values you derive.

I think that should clear up the question.

The one issue of Wilson’s letter is the “retraceableness” of using the Trimble units without explaining how you derived the bearings. If you’re going to base your bearings on that technique you need to have the Lat and Long stated on the map along with the mapping angle at the point since the software is essentially taking the theta angle at the “here” point and applying it to the bearings. It’s a GPS derived Geodetic bearing verses a Polaris or Sun shot derived Geodetic bearing which will have very little difference. If you survey is over 10,000 feet you may want to apply the Laplace Corrections.

You would still need to at least let the retracing surveyor know how you derived it, with information he could verify. By given him the Lat and Long and theta, he could even calculate it himself when retracing you mapping.

Please note:

There is about 1 degree difference between Grid bearings (CCS83, NAD83, WGS84) and geodetic bearings (LAT, LONG, SUN, POLARIS, SIDERAL).

The convergence angle (theta/gamma) between grid north and geodetic north varies with the longitude either east or west of the central meridian for the CCS zone of the point being determined. If you're on the central meridian, the difference between the two is zero. The magnitude of the angle increases as you head either east or west, but is negative in the westerly direction. (grid =true minus theta) Theta is negative west of the central meridian.

It's about a degree here in Santa Barbara.

It's true that the new State Code won't accept "autonomous" positions for Lat and Longs. So you're right on that point. However, explaining how you derived the bearings on the map will take some of the mystery out of it. It's about trust. Can the retracing surveyor trust that your bearings are "True" and not in error? Maybe it doesn't matter because they find two points. But it's hard to clarify your map when you're dead and gone. I'll have to try and pull this one off sometime.

On plus side I learned something: If you don't plug in a datum and projection, Trimble will convert the bearings to "True".

My question is: If there is no projection and no datum, what are the distances being reported?

The convergence angle at Santa Barbara is approximately a negative 58 minutes.

**CCS Zone 5**

Now I'm really confused.
Is there really one degree difference between the wgs84 bearings that are used by GPS ? if that's the case, then I'm wrong. also, KW is wrong on his basis of bearings note.

in 1995, I was told by Trimble, that GPS bearings were close to geodetic(true) plus or minus approximately 3"

The way I'm now seeing it, is that wgs84 is a modified cylinder around the earth, rather than a sphere. If it is a cylinder, then the bearings will differ at each lat/lon point of origin.
If the "here" bearings really correct the theta angle at the observed lat/lon , I'm drinking Dr Pepper.

If the WGS84 grid bearings are cylindrical as possibly indicated, then each "here" position, has a different bearing and is therefor randon with relation only to that of the individual survey.

I kept saying minute and not degree. At least it was a big mistake. It is "about" a minute here in SB. If you in Zone 5 the Central Meridian is at 118:00 Long. Santa Barbara is plus or minus 119-50 so we see about a degree. Corpscon can tell you what you should expect to see in your area.

I spoke with a Trimble Tech last night and the software is program to work in "True" bearings when no projection or datum is set for the job settings at start up as in your first email, so you are right (so long as the user does input a projection or datum). However, as many of you know a line of constant bearing on the surface of the earth is curved, not strait. Apparently they only recommend using the “one point calibration with true azimuths” for small sites since they are using the theta/ mapping angle at that “base” point where you did the here command. In addition I asked what the distances are being reported since the constellations operate in WGS84. I have a feeling they are grid distances since how would it know how to convert to ground without a projection and thus the recommendation to only use this one point calibration for small sites?

If it's using the WGS84 projection, then does TDS scale and report the coordinates in "Ground" distances? In 800' you should see (around here) about .08' difference between Grid and Ground distances.

I posed this question because to obtain a "Ground" distance you need to either survey in ground or when using GPS have a projection to calculate it based on a mathematical model such as WGS84. It's all about the software. The difference between WGS84 and CCS83 is so minor, that it will not affect your results unless your doing large scale Geodetic or Control Surveys.

I think it's a question worth asking.

I always use least squares for my control and boundary work, so I usually try to have it on a CCS83 basis and then if I want to map on a "Ground" or Geodetic basis I can convert it in the office and have the field crew do a site calibration. By doing this I eliminate erroneous measurements and know what type of error ellipses I'm dealing with.

I figure that it was a "plane" system but needed to make sure, since many of the project I work on are in excess of 10,000 feet (Rancho's, PLSS ect.) which is one reason why I keep it on CCS83 for convenience sake until the actual mapping stage. However, I can see the power of working in Geodetic bearings, if I was out there myself and stuck in the office where I could survey and set monuments at the same time! I use Star*Net more for quality control than to actually adjust the coordinates. It's also useful for combining GPS and conventional data when the crew aren't on a basis from the start.

Since WGS84 is a world-wide datum, and GPS satellites provide global coverage, it wouldn't matter whether you set up your receivers in California, Florida, Australia, or China, you should still be able to readily determine latitude and longitude to the sub-second level and a precise geodetic azimuth, without any need for a differential correction, local grid projection, or convergence angle. The lat and long will be precise as determined by the receivers, just not as accurate as can be determined with a differential correction to a known point.

Again, the terms "geodetic" and "astronomic" are not synonymous and their differences should always be considered when performing precise work. The "arc to chord" correction, also known as the "second term", used for determining the precise angular relationship between projected grid and geodetic lines is usually ignored for lines less than five mile in length. Laplace factors, however, should not be so readily ignored. According to NGS mapping of Laplace factors, you could see differences of up to 45" of angle in the western United States.

During my surveying career, I've met several surveyors, that due to their work locations within CCS zones, or maybe because of ignorance, assume that ground vs. grid distances are insignificant and therefore not a concern. If you're lucky enough to work near sea level and close to a standard parallel for the CCS zone, then the combined grid factor may be so close to 1.000000000 that you can ignore it. Same for the convergence angle. If you're real close to the central meridian for the zone, the difference between the grid azimuth and geodetic azimuth may only be a few seconds. Just always be aware that that's not always the case.

As far as lines of constant bearing being curved, I believe that the one exception to this is for lines of longitude, which represent geodetic or true north. Your back azimuth and foreward azimuth will be identical anywhere along a meridional line. Not so for any other line of constant bearing due to convergence at the poles of the meridians.

Geodetic and True north are the same - sort of. For this conversation they are, but the reality is that a SUN shot, Polaris and sidereal observation will be minutely different that the azimuth as determined from the satellite constellation because the satellites are on an earth-based orbit to determine the North Pole and celestial observations are not gravity based and affected by wobble and other gravity fluctuations. This is where the "Laplace" correction comes into play. I think for the purposes of most surveys, Geodetic = True.

A typical job I work on is 100-3000 or more acres, so I like to get a network set up static and then use RTK in a known datum (CCS83). Since the difference is about .01' per 100 and about a degree (rule of thumb here, so I don't need it down to the minute since we’re simple having a discussion and a surveyor should do this himself, for his area is he needs to) in the Santa Barbara CA you can recognize it pretty easily. I've had to do Laplace corrections too for the larger sites.

Since I spoke with Trimble myself, I learned something new- you can work in "True" bearings (as derived from the WGS84 constellation at a single point). However, I'm also wondering if they are reporting the distance in "ground". If so, then it's another tool in the toolbox as far as I'm concerned. I put in a question, but haven't heard from them yet.

Truth be told, I'm probably in way over my head on this topic. I think I'll gracefully bow out and defer any further discussion to the experts. I know there's some out there that are probably pretty amused by now with my ramblings.

"Santa Barbara is plus or minus 119-50 so we see about a degree. Corpscon can tell you what you should expect to see in your area."

Eric, could you please confirm that Corpscon can be used to compute convergence angles?

thanks

Yon don't need it, it's found in the Tables you can buy from CLSA but yes, CORPSCON will do it for you if you input the LAT LONG, N E of a point. It may be off a little without factoring the laplace correction however.

Thanks.

I found the routine on the 6.0 version. I'm more familiar with the programs in the NGS Geodetic Toolkit. I think Corpscon is designed around some of these programs, like Nadcon and Vertcon.