Examining the Discrepancy between Forecast and Observed Liquefaction from the 2015 Gorkha,Nepal,Earthquakes

Abstract

Many ground failures resulted from the 2015 Nepal earthquake sequence, including landslides, rockfalls, liquefactions, and cyclic failures. And whereas the amount and extent of landsliding were relatively consistent with predictions for a M_w 7.8 main shock, the amount and extent of liquefaction were not. We present a summary of liquefaction field observations that we made as part of the Geotechnical Extreme Events Reconnaissance (GEER) investigations. The liquefaction that did occur in the Kathmandu Valley was limited in its spatial extent, and the postliquefaction deformations were small. Prior earthquakes in this region have been reported to have caused greater liquefaction-related failures, and liquefaction hazard–mapping studies predicted widespread liquefaction hazard from an event of this size. We explore two possible reasons at the regional scale for the limited liquefaction from this earthquake sequence: drawdown of the groundwater table and high near-surface shear wave velocity. Our study finds that pumping has depressed the groundwater table across the Kathmandu Valley by 13–40 m since 1980, thereby decreasing the amount of near-surface liquefiable material and increasing the nonliquefiable “crust” layer. The regional slope-based V _S30 for the valley is on average higher than that for liquefaction sites in a global database of observed liquefaction. A global geospatial model for liquefaction occurrence shows low liquefaction potential in the Kathmandu Valley consistent with the observed patterns.