Atmospheric rivers (ARs) account for many of the costliest flood events in western U.S. maritime mountains. The flood potential of ARs is amplified when an antecedent snowpack is present. The New Year’s flood event of 1997 was representative of an AR-induced rain-on-snow event and remains the costliest in California’s history. The compounding influence of saturated soils, widespread rainfall, and abrupt snowmelt led to widespread inundation. AR-induced rain-on-snow events are projected to occur more frequently in a warmer world. Yet, the sensitivity of the rain-on-snow flood drivers to warming remains understudied, particularly at the spatiotemporal scales needed by water managers. Here, we leverage the regionally refined mesh capabilities of the Energy Exascale Earth System Model (RRM-E3SM) to recreate the New Year’s flood event of 1997. We simulate the 1997 flood event at grid spacings of 14km, 7km, and 3.5km, with forecast lead times of up to 4 days, and across six different warming levels. We then identify the sensitivity of various aspects of the 1997 flood event hydrometeorology to resolution, forecast lead time, and warming level. Specifically, we describe precipitation intensity, efficiency, rain-snow partitioning, and cold content sensitivities and how runoff characteristics respond to these factors. We also show how certain aspects of the flood event change when viewed from a storm total versus sub-hourly perspective. We aim to provide water managers with insights on how warming-induced changes to AR precipitation characteristics and snowpack dynamics could interact to amplify or diminish runoff potentials and guide preemptive, rather than reactive, climate adaptation.