Forced and internal variability of tropical cyclone track density in the western North Pacific

Forced interannual-to-decadal variability of annual tropical cyclone (TC) track density during 1979-2008 in the western North Pacific (WNP) is studied using TC tracks from observations and simulations by a 25-km-resolution version of the GFDL High-Resolution Atmospheric Model (HiRAM) that is forced by observed sea surface temperatures (SSTs). The model results generally agree with observations, indicating that much of the interannual-decadal variability is due to SST. The variability among different member runs of the HiRAM gives an estimate of internal variability.

The decadal variability is dominated by two modes in both observations and HiRAM simulations: a nearly basin-wide mode, and a dipole mode between the subtropics and lower latitudes. The former mode, which peaks in early 1990s, closely links to the variations in the annual TC numbers and might be driven by the decadal variations in SSTs over the northern off-equatorial tropical central Pacific. The latter mode shows a phase shift around mid-1990s and might be associated with the Atlantic Multidecadal Oscillation. On interannual timescales, HiRAM simulates two distinct modes. One is a basin-wide mode in association with SST anomalies of opposite signs respectively located in the tropical central Pacific and eastern Indian Ocean. The other mode exhibits an oscillation between the southeast and northwest quadrants of the NWP, and is linked to the conventional ENSO conditions. In observations these two modes are not well separated.

We further explore the seasonal evolution of the ENSO effect on TC activity via a joint EOF analysis using TC track density of consecutive seasons, which is more objective than the previously widely used correlation, regression, and composite analyses based on various ENSO indices. In observations, two types of ENSO stand out: a hybrid central Pacific and eastern Pacific (EP) ENSO, and a conventional EP ENSO. They differ in many aspects, including location and amplitude of maximum SST anomalies in the tropical Pacific, and pace of the decay. These, together with accompanied differences in the Indian Ocean SST features, generate different patterns of anomalies in atmospheric circulation, and thereby lead to different kinds of spatial and seasonal variations in TC track density. HiRAM reproduces these observed characteristics reasonably well.

The internal variability in the NWP TC track density is then examined using the ensemble simulations from HiRAM. On both the annual and seasonal basis, the internal variability is identified to be weakest over the TC main development region and is very strong in the South China Sea and along the East Coast of East Asia. And the randomness is larger during the peak and late TC seasons (July-September and October-December, respectively) than in the early season (April-June). The related internal variability in large-scale atmospheric fields will also be discussed.