Evaluation of Spatial Interpolation for RQD at Different Depths

Rock Quality Designation (RQD) of a site is a very important engineering property from geotechnical considerations. RQD is very useful in identifying potential problems related to bearing capacity, settlement, slope stability problems and also serves as warning indicator of low quality rock zones that requires greater scrutiny. However, due to time and cost considerations, geotechnical investigations are carried out at larger spacing which calls for appropriate techniques for estimation of data at intermediate locations for appropriate judgements. An earlier study reported IDW (Inverse distance weighing method) to be appropriate methodology for 20 m level below ground level at the site. It was deemed necessary to examine whether similar technique would be applicable for other layers also. This paper deals with estimation of RQD values at intermediate locations from an available coarse grid data using various spatial interpolation techniques for different depths. Geo-statistical methods selected for the present study are K nearest neighbour (KNN) mean method, Inverse Distance Weighing (IDW) method and Trend Surface Analysis (TSA). Spatial interpolation has performed at depths 20, 25 and 29 m below ground level. IDW and KNN methods estimated the RQD values at all depths with good accuracy followed by TSA method. It is observed that site-specific parameters obtained for best performance are different for different depths. It is necessary to analyse the data at each level and perform spatial interpolation to obtain the site-specific parameters to obtain best results.