Sensitivity of Northern Hemisphere Extratropical Cyclone Properties to Atmospheric Resolution in the GFDL SPEAR Model

Abstract Extratropical cyclones (ETCs) significantly influence midlatitude weather and climate through their roles in transporting heat, moisture, and momentum. Accurately simulating ETCs in global climate models (GCMs) is essential for reliable weather forecasts and climate predictions/projections. Previous studies have shown that increasing the horizontal resolution of GCMs can enhance the representation of ETC characteristics. However, the impact of atmospheric resolution finer than 100 km on ETC genesis and frequency, especially when considering the effects of post-processing spectral truncation during cyclone tracking (e.g., T42 or T63), remains underexplored. This study addresses these gaps by examining the sensitivity of Northern Hemisphere ETCs to atmospheric resolutions of 100 km, 50 km, and 25 km using the fully coupled GFDL SPEAR model. Cyclone tracking is performed at spectral truncations of T42, T63, and T106 to assess the effects of spectral smoothing. Results reveal that increasing atmospheric resolution has minimal impact on the frequency of large-sized ETCs but leads to a substantial increase in the frequency of small-sized ETCs, particularly over the North Pacific and the North Atlantic. The size-dependent sensitivity of ETCs to atmospheric resolution is the most pronounced at T106, and further analyses suggest that the genesis of these small-sized ETCs are highly impacted by diabatic heating. Comparisons with various reanalysis datasets suggest that high-resolution models are likely better at capturing small-sized ETCs rather than overestimating their frequency. These findings highlight the critical importance of considering size-dependent sensitivities when interpreting ETC biases in high-resolution model outputs compared to reanalysis data and underscore the potential role of diabatic heating in genesis of small-sized ETCs.