Detection of Height Dependent Error in Precise Leveling on a Mountain Slope

Abstract

In a precise leveling survey, height dependent errors often arise on a mountain slope. They consist of rod calibration error and refraction error. In order to detect these errors, it is necessary to carry out the survey, first, using the leveling staves the scale bands of which have known and small temperature expansion coefficient (hereafter, TEC) and, second, measuring the air temperature gradient near the ground surface at each survey point for the correction of refraction. For this purpose, two sets of staves were made from super-invar and new superinvar. The original design of which was done by one of the authors (TANAKA). This was carried out under the cooperative research of the Research Institute of Electric and Magnetic Alloys and Sokkisha. Co. Ltd. The bands of staves have the TEC of 0.55 × 10^-6/°C and -0.1 × 10^-6/°C for super-invar and new super-invar, respectively. Their graduation errors are within 15 μm/3 m. In order to test the new staves, comparison surveys were carried out two times, each time the survey was done by the different two teams, near Ito, eastern part of Izu Peninsula. In the survey, one test team used the new staves and the other team ordinary ones. As the result, the height dependent correction amounts to 4 mm/100 m was detected. On the other hand, the height dependent error 6.7 mm/100 m was detected by treating the recent survey results statistically. In order to understand the difference of height dependent error in the test result and the statistical one, it is necessary to make more research on the refraction together with the effect of wind turbulence. It is concluded that in order to reduce the correction value less than 1 mm/100 m, the accurate control of staves is the most important in the precise leveling. It is also possible to eliminate the refraction error by keeping the sight length short.