Low snow levels over the past few decades and predicted in the future have prompted the emerging research of snow droughts as snow is critical for water security and management. There are many data products and model simulations of snow water equivalent and snow cover available, but no systematic data-model comparisons and understanding of the drivers of snow drought changes in the past. We analyzed snow drought events using standardized snow water equivalent index calculated from the monthly results of four numerical experiments using the E3SM Land Model (ELM) during 1980-2014 and compared with those calculated from the ERA5-Land data. The numerical experiments were conducted with atmospheric forcings from two quasi-observed forcings (GSWP3 and GSWP3-W5E5), and two forcings from AMIP-like simulations using E3SM. GSWP3-W5E5 contains data from the GSWP3 dataset and the W5E5 global meteorological forcing data processed for Inter-Sectoral Impact Model Intercomparison Project. The two AMIP-like E3SM configurations differ in whether plant hydraulics (PHS) feedback is simulated or not. All data show increased intensity and frequency of snow droughts in between 2000 and 2014 compared to the years before. In general, simulations driven by AMIP-like forcings produce snow droughts during the 35 simulated years that are more comparable to the drought events in ERA5-Land than simulations driven by GSWP3 or GWSP3-W5E5, especially for the low-to-mid altitude area in Asia during drought years from 2005 to 2008. The drought events from all simulations are low in frequency in Western Europe compared to those from ERA5-Land from 1992 to 1998. Including PHS feedback resulted in more drought events in Europe. Our analysis show that the prediction of drought events is subject to the uncertainties in the driving force and E3SM has the potential to be used as a useful tool for predicting low snowpack and preparation of their impact in the future. Drivers of snow drought changes and data and model discrepancies will be discussed.