The use of surface interpolation methods for landslides monitoring

Abstract

The subject of this study is selection of mathematical description (interpolation method) of landslides surface for the purpose of periodic control of terrain movements. Monitoring of landslides, considered in this work, concerns points of periodic GPS or tacheometric measurements carried out in a relatively regular grid. For economic reasons, the grid usually consists of limited group of about several dozen of points that do not allow for an exact description of the terrain surface, nevertheless a general movement of a landslide can be detected (observed). Replacement of discrete set of points to a continuous surface model allows for a better assessment of the phenomenon especially within the scope of its main directions of activity. The choice of best performing interpolation method in order to construct a reliable numerical terrain model was performed through tests carried out on complex model surfaces. These model surfaces encompassed various configurations of terrain relief. In this study four interpolation methods i.e. splines, kriging triangulation with linear interpolation and IDW are compared. Tests on model surfaces distinguished the method of splines as the best performing one as well as the method of kriging which is suitable in some instances. The selected method of splines was then applied to periodically monitored landslide in Milówka in the south of Poland. Interpolation by means of splines allowed for observing progressive movements of the landslide surface with accuracy up to several millimeters. This confirms the usefulness and reliability of this method in landslides monitoring. Specialized field: Surveying & Geo-Spatial Engineering: Deformation Surveying.