Classification of Limestone Rock Masses Using Laboratory and Field P-wave Velocity by ArcGIS Fuzzy Overlay (AFO) (Case Study: Five Dam Sites in Zagros Mountains, Western Iran)

Abstract

In this study, 18 seismic tomography profiles in limestone rock masses with a total length of 2450 meters were studied to find the relationship between P-wave velocity in field and laboratory (VPField and VPLab) and rock mass classifications (Q and Qsrm) in the Zagros mountains, western Iran. The results showed that the Q rock mass classification system and its modified system for sedimentary rock masses (Qsrm) were closely related to Vp parameters. In addition to the VPField, KP (VPField/VPLab) also showed a significant relationship with Q and Qsrm. The best multivariate equation between VPField and KP with Qsrm (R2 = 0.70) was more reliable and had less error than the best between VPField and KP with Q (R2 = 0.55). The reason for this was the status of the voids and some other rock mass properties such as bedding and structure of rock mass in calculating the Qsrm, which were neglected in calculating the Q index. Certainly, voids and many rock mass properties in the limestone were very effective in the amount of KP and VPField. Also, the development of models of ArcGIS fuzzy overlay (AFO) methods for prediction of Qsrm showed very interesting results because it was able to show stronger accuracy (R2 = 0.89). In order to check the accuracy and errors of the obtained models, “Error%”, “Mean Absolute Percentage Error”, “Mean Absolute Deviation” and “Root Mean Square Error” indicators were used which showed the high-efficiency of the AFO model in estimation the Qsrm. According to the results of this study, it was proposed: in the studied calcareous rock masses, Qsrm classification can be a suitable substitute for Q classification.