There is a misunderstanding shall be clarified first that precision of resection is the highest because there is no alignment error of survey station and instrument height, some people like to take CPIII measurement in high-speed railway as an example.

For CPIII measurement in high-speed railway, it is conducted in a very small surveying range. Besides, the know points for resection are far more than two points.  The purpose for CPIII measurement in high-speed railway is to make precision measurement. What matters for it is precision. It is the precision at 1mm or even lower. In this case, resection is the best method.

However, in common daily survey operation, there is no need to pursuit for that precision. For example, resection can directly get the station height (i.e. the sight line of the instrument), to eliminate the survey error of instrument height. However, under normal circumstances, the instrument height measurement error is at most 3mm. Do you really care about the 3mm error in your job?

In common daily survey, what we seek is to ensure the necessary accuracy in the case of field works as reliable as possible. Generally, it is enough to reach the precision at 1cm. In most cases, reliability is far more important than precision. Possibility of large errors will be increased by resection with just only two known points in common daily survey operation, which is inadvisable. Besides, what shall also be specially paid attention to during resection is the selection of geometric graphics.

Main risks of resection are as follows:

①Large data errors will occur while operating the instrument. In the process of setting known points, saving coordinates of survey station and so on, it is very easy to make mistakes. For example, coordinates of survey station shall be retrieved when it is needed to continue to use transition points after observation.

②There are gross errors caused by unreasonable position of survey station. For example, survey station and two control points are placed almost on the same line, resulting that it is not easy to find gross errors occurred. There are many same problems as this in the daily measurement, but most of them can't be understood why gross errors occur.

③For another example, the instrument is moved during observation. It is needed to recenter and reorient, but the point being observed is not marked. All these pitfalls are possible.

How to prevent from the above risks?

First, the most traditional way is to set survey station at a known point, use transition points and then go forward to observe.

Second, do not use resection to avoid the above gross errors, except for precision measurement like CPIII survey. In this way, possible errors can be predicted in advance, allowing to be free from gross errors. This is far more important than improving precision by a few millimeters.Not all surveyors can understand the potential risks of resection, not everyone can understand the principle of it, unless all your surveyors are trained or well-educated.

If it is required for using resection, please kindly find the graphics shown as follows.

As shown in the figure, the strength of No.1-3 graphics is enough, and the precision is reliable. However, the errors of No.4-6 are large. It is difficult to improve the precision no matter what kind of observation you do in these cases. Among the graphics No.4-6, the horizontal error of No.4 is large while there are large vertical errors for No.5-6. Generally, it shall be avoided to use such kinds of graphics.

How to do when it must use resection in the condition required for high precision, but the graphic condition is poor? Either encrypt the control points or improve the angle precision by returning to measure for multiple times, observation on left and right angles, repeated setting station and other methods (the distance precision of TS is enough to meet your needs, and the major errors are from angle measurement).

But it is not meant to completely exclude the resection, it shall be operated depends on the actual situation. Based on real needs, it shall be confirmed whether it aims to make sure the accuracy of stakeout, extreme precision or solve the issue for lack of intervisibility among control points.

There are advantages of rear resection. But it also shall know about the source of its errors for better use, especially in deformation monitoring, during which there is no methods with better precision than resection-The premise is that it needs to be used correctly.

________________________________________________________________________________________________