A Spatial Framework for Comparing GCM-Generated Tropical Cyclone Tracks

Following recent improvements in model resolution and physics, coupled ocean-atmospheric general circulation models (GCMs) are able to generate vorticies resembling actual tropical cyclones (TCs). We can therefore now obtain simulated TC tracks from the output of these high resolution GCMs. Various clustering and spatial kernel techniques, which rely on numerous free parameters, have been employed for track comparisons. Here we instead propose a spatial framework that allows for flexible comparison of tracks from observations and GCMs. The framework is a tessellation of equal-area hexagons that enables comparisons between track data and other TC attributes. We apply this framework to compare GCM-generated TC tracks from the GFDL HiRAM-C180 with best-track (HURDAT) observations over the period 1981-2010. This framework will allow a unique and unprecedented look at how GCMs are able to capture the local and regional hurricane climate processes on seasonal to decadal time scales.