Tropical cyclones in a T159 resolution global climate model: comparison with observations and re-analyses

Abstract

Tropical cyclones have been investigated in a T159 version of the MPI ECHAM5 climate model using a novel technique to diagnose the evolution of the three-dimensional vorticity structure of tropical cyclones, including their full life cycle from weak initial vortices to their possible extra-tropical transition. Results have been compared with re-analyses [the European Centre for Medium-Range Weather Forecasts (ECMWF) 40-yr Re-analysis (ERA40) and Japanese 25 yr re-analysis (JRA25)] and observed tropical storms during the period 1978–1999 for the Northern Hemisphere. There is no indication of any trend in the number or intensity of tropical storms during this period in ECHAM5 or in re-analyses but there are distinct inter-annual variations. The storms simulated by ECHAM5 are realistic both in space and time, but the model and even more so the re-analyses, underestimate the intensities of the most intense storms (in terms of their maximum wind speeds). There is an indication of a response to El Ni˜no-Southern Oscillation (ENSO) with a smaller number of Atlantic storms during El Ni˜no in agreement with previous studies. The global divergence circulation responds to El Ni˜no by setting up a large-scale convergence flow, with the centre over the central Pacific with enhanced subsidence over the tropical Atlantic. At the same time there is an increase in the vertical wind shear in the region of the tropical Atlantic where tropical storms normally develop. There is a good correspondence between the model and ERA40 except that the divergence circulation is somewhat stronger in the model. The model underestimates storms in the Atlantic but tends to overestimate them in theWestern Pacific and in the North Indian Ocean. It is suggested that the overestimation of storms in the Pacific by the model is related to an overly strong response to the tropical Pacific sea surface temperature (SST) anomalies. The overestimation in the North Indian Ocean is likely to be due to an over prediction in the intensity of monsoon depressions, which are then classified as intense tropical storms. Nevertheless, overall results are encouraging and will further contribute to increased confidence in simulating intense tropical storms with high-resolution climate models.