Introduction

Dynamical Downscaling of Western North Pacific Tropical Cyclone Genesis

by the IPRC Regional Climate model

Yuqing Wang and Ruifen Zhan

¹International Pacific Research Center, University of Hawaii at Manoa

Honolulu, HI 96822, USA

Abstract

Dynamical downscaling with a regional climate model (RCM) has been widely used for climate change assessment, seasonal predictions, understanding of climate processes and regional climate predictability (see the review by Wang et al. 2004). By this approach, a high-resolution limited area model is driven by relatively coarse resolution reanalysis or output of a coarse resolution global model to produce more detailed regional features. This has been approved to be particularly useful for the study of extreme climate events, including tropical cyclones. There have been several studies using RCMs to investigate the possible impact of global warming on the activity of tropical cyclones in different ocean basins. Knutson et al. (2007) demonstrated how well a nested regional atmospheric model driven by the reanalysis reproduced the recent multidecadal increase of Atlantic hurricane activity. However, the hurricane counts in the model were very sensitive to how strong the large-scale nudging in the interior of their model was.

In this talk, following a brief review in the dynamical downscaling of tropical cyclone genesis, the performance of the regional atmospheric model (IRAM) developed at the International Pacific Research Center (IPRC) in reproducing the tropical cyclone genesis over the Northwest Pacific has been evaluated. The NCEP/NCAR reanalysis is used to provide both initial and lateral boundary conditions to the IRAM, which was run at a horizontal resolution of 0.3^o lat/lon. The weekly mean Reynolds SST was used as the lower boundary conditions for the model over the ocean. Simulations for 17 years from 1990 to 2006 are analyzed in the study. The model was initialized at June 1 and integrated continuously through November 30 for each year without the use of any interior nudging. The model TCs are located and tracked using an objective algorithm previously developed by Stowasser et al. (2007) for the IRAM. The results show that the model reproduced not only the mean seasonal cycle but also the interannual variability of tropical cyclones in the Northwest Pacific realistically. The correlation coefficients between the modeled and observed annual tropical cyclone counts and the power dissipation index reach 0.82 and 0.74, respectively. The correlation coefficient between the modeled and observed seasonal cycle reaches 0.92. Further, the model also reproduced the spatial distributions of genesis locations and frequency occurrence reasonably well. Further experiments for the Northwest Pacific TC genesis in response to sea surface temperature anomalies over the eastern Indian Ocean show that the model can reproduce the observed response quite well. Therefore, our results demonstrate that given accurate large-scale circulations by a coupled global model, the IRAM can be used to provide skillful dynamical downscaling for seasonal prediction and climate change projection of tropical cyclone genesis over the Northwest Pacific.