Earth's Largest Land Surveyor Community
I would like to know a methodology for traversing between two known points that cannot see between the two points. That is you cannot occupy one point and see the other.
At Ohare airport in Chicago I have several known monuments. I would like to traverse between two points to establish control between the monuments. However, if I start on one point I will not have a known back sight. Also, the two points are not within sight of each other. Several turns will be required.
How do I go about this with no backsight?
This Content Originally Published by a land surveyor to Land Surveyors United Network
It just depends, on what you can see and use. If the two points can be seen by doing a resection then you started a new problem. Although a 2 point resection is possible to have high accuracies,it is not wise to use them. You need at lease three known points or more, to have a resection with redundancy. So a resection may not be the solution,only 3 or more points was my intent when I suggested using one. One could also set up a Triangle network later after everything has been worked out so to have 2 or more future BackSites from a known point and also reference sighting Azimuths to radio towers or other objects etc.
A 2 point resection may start you but it will not end there, because you have no redundancies, in this solution of a Total Station survey. Now you have to prove or adjust this solution the same as before,
Square one all over again. still need a closed traverse and adjustments.
A resection will not work because of line of sight obstructions.
Using GPS is not ideal due to tall buildings and trees. I did set a gps point to use an azimuth to use for a back sight while occupying one of the known monuments. However, I do not feel that this is very accurate.
This depends on what the data on the monuments is. Was this data conventional or was
GPS used. If not, this GPS Azimuth does not relate to the Azimuth between the known monuments anyway it is no more than a assumed Azimuth..This is like any other survey were you traverse with assumed coordinates and tie into known points. Tying both points if that is all you have, you said you occupied one using the known coordinates then when you tie the other you can inverse to see the different's in bearing and distance,then all you have to do is roll your knew points to the old bearings between your known 2 points. Closing a traverse is just to make sure of everything.It is hard for people to see this without looking at it. The bearing between your known points is what matters to your rotation. If you start with known coordinates on one of your points then all you will have to do is rotate your new points from that point the one you started on, that is your rotation pt.
So you know the coords of the two points you are traversing between right? You can calc your known bearing between these two points. Just set up on a random point between them, backsight the first and then traverse however you need to get to the second one. Then inverse between your first and last field shot the match up to your known and compare the distances, If they check good enough for you then just do a hermite transform in your data collector. It should be really simple.. or did I miss something?
Let me know,
Setup on 1st point and perform a solar observation to a reference mark, reduce it to true or grid azimuth, do the same for the second point, setup on 1st point backsighting the reference mark and traverse to new points and close on your second point and closing angle on your second reference azimuth mark, .. adjust angles if needed,.. and adjust by compass rule..
easy old school and done...
this is how things were done before GPS
DO a ray trace: occupy an arbitrary point, set an arbitrary orientation to one of the monuments, traverse through a set of points to the 2nd monument, adjust using the datum bearing and distance.
In that case, considering the nowadays accuracy in angles and distances, better to not even adjust angles but to loosely integrate the solar azimuths at both ends. That can be handled by least squares adjustment considering the respective standard deviations (angles, distances, azimuths) that will be anyway "transformed" into corresponding weights.
Having said that, solar azimuth requests some experience and a special prism (Roeloff) ... time is not more an issue. The accuracy of solar azimuth can be on the range of +/- 0,0005 gons assuming you know also your position. Then we are back on GNSS.
That's rather simple indeed ... You start your computation fixing the coordinate of the first starting point (known coordinate) and an approximate orientation (bearing or azimuth) then you propagate the coordinates till the ending point. When you have your coordinates you can perform a 2D transformation (Helmert) and derive the DX, DY, Orientation and Scale, then transform all your stationing points. Normally the scale will indicate you how much your quality will be. Now to introduce some redundancy, you can double your station points ... Instead of measuring A - 1, 2, 3, 4 - B (assuming A and B are your control points) you can measure A - 1 and 1b, then 2 and measure to 3 and back to 1 and 1b, then 3 and measure back to 2b etc, ... It's just a question to multiply your measuring data ... Good luck and don't hesitate to contact us on : [email protected]
GPS Network will not "arrange" the traverse between 2 known points, it will provide the control points but you have situation where you cannot survey by GPS/GNSS all the traverse points. We had the vase for instance in Hong Kong in a very urbanized area with tall buildings. So using Total Station for densifying a network will still be the solution. Now, regarding my post above, you can adjust a traverse between 2 known points without having orientation at both end. You just have to consider the traverse as "line" that must be rotated and translated on the known points. As you will have to check in that case, we suggest to multiply the stations a bit like we are doing for leveling lines. One station on 2, should be duplicated and not necessarily far away, few meters away is already good.