AbstractEstimated extreme precipitation and floods provide crucial information for both the design of large hydraulic structures and water resources management. This study aims to estimate, in a series of numerical experiments, the probable maximum precipitation associated with extreme rain events over seven watersheds in the central/southern Sierra Nevada. To this end, a total of 249 extreme 72-h rainfall events, extracted from the historical reconstructions from 1852 to 2014, were maximized by means of the atmospheric boundary condition shifting (ABCS) method in meridional (north/south) and zonal (north/south and east/west) shift directions. Furthermore, optimization of relative humidity along atmospheric river (AR) corridors was applied to 20 significant storms, which were identified from the ABCS results. The meridional shifting exhibited a significant increase compared to historical conditions, with an additional slight increase by the zonal shifting method. Then, the relative humidity optimization exhibited a further significant increase in precipitation over the seven watersheds. The proposed method can generate numerous events at various shifting quantities and relative humidity intensities along the AR. The selected events can then be used as inputs for hydrologic simulations to estimate probable maximum flood (PMF) events.