Basic Traversing Standards

Prepare yourselves... as this might be the first of many, what may be viewed as dumb questions.

We're currently doing a large scale traverse (approximately 30,000 ft.) in preparation for a job.

Our current party chief... yes the pot-smoking one... is heading up the crew.  He believes that the maximum distance you can traverse between points is only limited by how far you can see, as long as your points are balanced. 

I believe that even with today's technology, it's still a good rule to limit yourself to 300' or less, regardless of how far you can see.

What say you, collection of survey Gods?  :)

I'm not one of those, but I'll chime in anyway.

For horizontal control, long sight distance is good.  Fewer setups mean fewer chances for blunders, centering error, and sighting errors.

For vertical, it is good to limit sight distances to about 300 ft.  Beyond that the refraction and curvature corrections get large (depending on distance squared) and won't cancel out unless the balancing of distances is especially good.

longer is better, and in traversing you can go 500'+......  as bill said, watch your verticals though...
otherwise I always try to make my foresights and backsights long...  I'm only liited by the pine trees and scrob oak, plus various other junk undergrowth.  Mine are usually less than 300'...unlessi get lucky!

I'll go with the pot smoker on this one, I like long sights in a traverse, the longer the better. Depending on equipment and proceedure, you can also get pretty good vertical too. But I like to run levels, if I think the vertical will be critical to the job, or any future work to be done from the control points.

If the traverse is for boundary only sure... as far as you can see

BUT if you are going to run elevation I would not go much further than 500' per jump

I would concur that longer sights the better.  No 600 ft on one side and 100 ft on other unless absolutely necessary .

Brace traverse by traversing across mid way.

Use tribrachs to interchange instrument and target (forced centering)

READ angles min 4 times on two faces unless have TS with compensating correction.

radu

While it is true that best results will be had with equalled back and forward sights, what you are able to do in the field is sometimes what you are presented with.

Check and adjust equipment (everything including tripods) as needed will be most helpful.

Double and record all angles and use well adjusted poles or tribrachs for back and forward sites. Check distance to backsite before turning ahead.

I have ran 12mi  traverse with excellent results that had sites that were from 300ft or less to 1,500ft or more from the same point. Sometimes because of terain or obstacles, you get what you are served and live with it.

If you have GPS, make some static setups to check your traverse against.

Hopefully you will be able to be able to make cross property closures.

I agree with the "longer is better" IF:

1) It is for horizontal control only.

2) It is STILL run relatively close to the boundary or property lines.  I don't know how many times I have followed crews that took that nice long shot out to a pasture to get out of the woods and completely missed that old fence embedded in the trees, that old hedge row, the ditch line, or even (gasp) and old family burial ground.  Quick and easy still doesn't relieve you of the responsibilty to check those boundary lines.

3) Correct field procedures are used (i.e. correctly adjusted equipment, multiple angles & distances, correct set up procedures, etc.).

Andy

1. if for control - longest visible distance
2. if for topographic survey - place points where you can observe the most number of points.

to make good closure for traverse, i always use gps for points that are open to the sky and close the traverse points on the gps points.
then i run a differential level loop to connect all the traverse points.

30,000' for traverse is about 6 miles. If you  were in a PLSS state one might assume you were doing the perimeter of 2 sections. If it is in Virginia the shape would be more irregular than a rectangle. Assuming all traverse shots were 500' you would be setting 60 traverse points. You will probably have more. In order to tigthen up such a large traverse, I would suggest 6 sets of GPS checks. In my opinion a GPS set to check for a traverse should be a set of 3 traverse points in series. That gives you a check of 3 coordinates, 2 lengths and 1 angle. Let's assume you will do 6 sets of 3 for 60 traverse points. Assume Pt 1 will be your overall site control and you have only 2 L1 GPS receivers, set on 1 and put the second on 2, then 3, then 2, then 3, moving it to 11, 12, 13, 11,12, 13 and do a third setup on 11 or 12, then one to 21-23, 31-33, 41-43, 51-53 and closing at 2 and 3. The same scheme can be used with RTK. If you have 3 receivers do 1,2 & 3 then move 2 receivers to 11, 12 then 12, 13, then continue to 21, 22 and 22, 23 etc. closing at 2, 3. The object is to make one of each set of 3 with the best skyview as the local control, with more observation time than the other 2, giving it more weight in your LS adjustment. once you have 2 sets you can adjust your traverse in segements to check for blunders, before the traverse is complete and you are well away from the project.

You can reduce the number of sets to 4 if you can do a cross traverse tie. There is no requirement to do exact increments, let your skyview dictate the locations. Say 1, 2 & 4  then 10, 12, 14 but keep them bunched to facilitate checking. Also if you have an internal azimuth point that can be observed from at least 10 scattered locations you can reduce by one set. I would say 1 internal azimuth point is equal to 3 external azimuth points unless you can observe each the externals for along at least 6 points along 1/2 the traverse length. It is a bonus to have azimuth points with recent coordinate values. A CORS within 12 miles can significantly reduce the time of L1 observations and CORS within 60 miles can be considered as control points for static L1/L2 GPS.

With nearby CORS you can make your GPS the control point and GPS your sets as you traverse about making the extra time in the field only for GPS setup and tear down. If you reduce your GPS to pairs instead of sets of 3 you reduce your checks by 1/2 with only a 1/3 reduction in GPS time. 
Do not hesitate to make a GPS point with excellent skyview if it is not on your traverse. A GPS point observable from 2 traverse points is as good as a point on line. If the offline point is matched with an online GPS point just tie it in tight with an angle from the backsight as well as the foresight. If you want to minimize the triple sets put them at the fartherest corners with pairs in between. Even a single interspersed GPS point helps but only a little compare to the time advantage and precision of sets.

Paul in PA 

The longer traverse lines and fewer stations may be an advantage to precision, however, you must consider the evidence of the boundary you may miss with longer distances. By long I would assume to mean thousands of feet. We need to traverse as near the property line as possibe in a boundary survey and the purpose of this guidance is that important evidence must be included in the work.

In work that I have done long sights have sometimes been compromised by scintillation and changes in the atmosphere in between the sights. Plan carefully and use GPS static positions for redundancy as well as well as redundant observations of traverse stations. Discussing methods for redundancy  would be better than whether to make the sights long or short. So I try to make the sights as long as possible without compromising the purpose of the survey.

See... this is why I suck at surveying.  :p

Anyway, I should have mentioned that we are carrying vertical through these points.  However, our LS also wants levels run through the whole job.  So I guess the vertical doesn't really matter.

I guess it was just so drilled into my head in school, that you should keep the distances between points "resonable", which turned out to be about 300', so that's what I've always gone with.  Personally, I have trouble even seeing further than that, with any kind of confidence.

That 300' thing you learned in school was most likely for levels, not total stations and traversing... I find 300' to be almost my maximum shot with the old autolevel... and that's a leica 32X level..

Have trouble seeing what at long distances?  You aren't trying to see 0.005 or 0.04 ft or some fixed size of error at any distance, you are trying to get an accurate angle.  If the focus in the telescope is working, your angular error of sighting should be constant at any distance.   So the fewer angular errors, centering errors, and instrument errors you have, the better the end-to-end angle and by D*tan(angle error) better end-to-end sideways distance error.

You could run a least squares on a made-up example with realistic values of centering and sighting/instrument angle errors to see the difference in the end result error ellipses.  That would tell you more than all the opinions posted here.

Minimum should be six inch tilting prism targets on tripods. On a job like this you want to use 4 tripod setups, with someone always moving a tripod ahead. The iman only worries about the instrument and the shots. That extra man can easily increase the work output by more than his incremental cost. Depending the crew capabilities you can switch out the grunt work. If the grunt is totally inexperienced let him do backsight pickup only. He then carries it forward to the current foresight position and lets an experienced crewman move it forward for the next foresight setup. 4 tripods requires a minimum 3 man crew. 2 men are maxed out with 3 tripods.

Carry vertical angles anyway, it is a great check on your overall traverse precision. That and GPS can quickly point out busts in your level work.

Paul in PA

6" tiliting prism's?   Mine are only 3" and they are plenty adequate...

I said 3" and actually they are 2.5"....  all my leica glass is 2.5"......on a rare occasion I have use the triple prism holder (twice in 30 years).

The peanut prisms that we use are 1".....
Maybe we're talking different things... my glass is 2.5", and all the prisms that I looked at 2.5" except for the peanut prisms.   you must be including the frame target, etc.. not just glass size...

ASCSurveyor:
See... this is why I suck at surveying.  :p

Anyway, I should have mentioned that we are carrying vertical through these points.  However, our LS also wants levels run through the whole job.  So I guess the vertical doesn't really matter.

I guess it was just so drilled into my head in school, that you should keep the distances between points "reasonable", which turned out to be about 300', so that's what I've always gone with.  Personally, I have trouble even seeing further than that, with any kind of confidence.
 

Don't be so down on yourself, a negative attitude will get you no where, fast.

Running a traverse around the campus isn't exactly like the real world, you will run into many different obstacles that don't exist in a lab setting.

Try to think outside the box and most of all, try to get along with the people you are working with.

Radar

A target may contain a prism, but for accuracy sake you align with the target, not the center of the prism.

You can somewhat improve angles with a candy cane on a straight prism, but accuracy on terrain requires a tilting prism that is aligned and tilted to the line of sight.

Paul in PA

My vote is "Long is strong." 

That is because I fear blunders more than I do error.

-David

I like to limit topo shots to about 500 feet and set up my control with that in mind. Usually about 700-800 ft between control points does it.

Try telling My TS with automatic prism recognition that it must sight the target  and it will give you the up you single digit finger!

It will also say to you that it does not even need to sight the centre of the prism as it has built in intelligence to compute the offset correction to the angle read to where the ray first hits the prism in order to save time for TS to precisely sight centre of reflector.....


RADU

For the science you will need to ask Mr Leica. All I know is that once beam finds the reflector and TS receives a return signal it is determining how far that beam is off from the centre of the reflector and computing delta to be applied to bearing reading in time it takes to meaure the distance.

RADU