Comparison of Arctic Properties in CESM2 Large Ensemble and ERA5

Even as global climate models have improved in their representation of Arctic processes, capturing observed trends in certain Arctic processes, such as sea ice and Arctic cyclones (ACs), continues to prove difficult. The release of the CESM2 Large Ensemble data provides the opportunity to re-examine how these processes are represented in models. We hypothesize that there will be improvements from previous generations of climate models but still lack precision when compared to observations.

There are three main avenues of investigation we focus on. First, we characterize ACs and Tropopause Polar Vortices (TPVs) in both the CESM2 LE and ERA5 reanalysis. This is accomplished by applying a watershed method tracking algorithm to the data, creating a dataset of AC and TPV cyclone tracks. We are then able to compare properties of those cyclones, such as genesis and lysis locations, lifetime, radius, and strength, to cyclones in reanalysis data and previous model studies generated by the same method. Secondly, we are interested in how well Very Rapid Sea Ice Loss Events (VRILEs) are captured by the model output. As VRILEs are significant sea ice loss events on the timescale of days, previous generations of global climate models have failed to capture them. Additionally, since VRILEs are typically associated with both an AC and TPV, how those features are represented in the model likely plays a role in the processes that result in VRILEs. Finally, we are interested in the time evolution of power spectra of sea ice extent in both CESM2 LE and observations.