Generally, errors cannot be totally eliminated, but they can be contained within acceptable tolerances. This requires you to use the prescribed methods and instruments and apply corrections established either mathematically or by experience. Some of the conditions that produce errors are listed below.

1. Instrument not properly adjusted. A small amount of residual error will always exist in any adjustment. For the more accurate surveys, the residual error can be minimized by using BS and FS balancing and, in trigonometric leveling, by taking direct and reverse (circle left and circle right) readings for the angles. 

2. Instrument not leveled properly. Unlike the residual adjustment error that will affect the readings one way consistently, this is a random or accidental error. It may affect the line of sight differently at each setup. This error can be minimized only by careful leveling each time the instrument is set up and by recentering the bubble before each reading.

3. Telescope not focused properly. Misfocusing and parallax in the eyepiece create accidental errors that cannot be corrected. The only way to avoid or minimize this error is to take care to focus properly at each setup. The instrumentman should check and clear parallax before the first sighting and should not readjust it until all sightings from the setup are complete.

4. Rod improperly plumbed. This error is caused by a rodman who does not pay attention to his work. The instrumentman can call attention to plumbing if it is at a right angle to his line of sight, but he cannot see it in the direction of line of sight. The use of a rod level or waving the rod will avoid this error.

5. Unstable object used for a TP. The rodman causes this error by selecting a poor point of support, such as loose rocks or soft ground. As the rod is turned between sights, the weight of the rod can shift a loose rock or sink into soft ground. The elevation of the TP as used for the next BS can change appreciably from the value that had been computed from the previous FS. This error can be avoided by using the turning pin or pedestal when the ground does not present solid points.

6. Rod length erroneous. This error results in either too long or too short rod readings at each point. In a survey predominately over slopes, this error will accumulate. The rod length should be checked with a steel tape at intervals to locate this error.

7. Unbalanced BS and FS distances. The unbalanced distances do not cause the error. It is caused by the effect on the line of sight from residual adjustment and leveling errors and the effect of curvature and refraction errors. Readings you take at a long distance will have a greater error than those at a short distance. This unbalance may not be critical on one setup but can be compounded into a considerable error if the unbalance continues over several setups. By balancing the sight distances at each instrument setup, if possible, and the sums of the BS and FS distances at every opportunity, you will keep these errors to a minimum.

8. Earths curvature. This produces an error only on unbalanced sights in leveling. When the BS distances are constantly greater than FS distances, or vice versa, a greater systematic error results, especially when the sights are long.

To eliminate this error, you must maintain a balanced sight distance in every BS and FS reading, not just their sum total between BMs (the error varies directly as the square of the distance from the instrument to the rod).

9. Atmospheric refraction. This error also varies as the square of the distance but opposite in sign ( + or ) to that caused by the earths curvature. The effect of atmospheric refraction is only one-seventh of that caused by the earths curvature. In first- and second-order leveling, the effect of refraction is minimized by taking the BS and FS readings in quick succession and avoiding readings near the ground. (They should be taken at least 2 ft from the ground.)

10. Variation in temperature. If a portion of the telescope is shaded and some parts are exposed to the suns rays, it produces some warping effect on the instrument that may affect its line of sight. This effect is negligible in ordinary leveling; but in leveling of higher precision, this effect may produce appreciable error. This is one of the reasons why surveyors use an umbrella to shield the instrument when doing more refined work.