Intensified Arabian Sea tropical storms

Tropical cyclones over the Arabian Sea in the pre-monsoon season (May–June) have intensified since 1997 (ref. 1, Fig. 1a) owing to significant reductions in storm-ambient vertical wind shear (VWS) in the troposphere; these reductions have decreased on average by about 3m s from the pre-1997 epoch (1979–1997) to the recent epoch (1998–2010). The authors attribute the reduction of premonsoon VWS to the dimming effects of increased anthropogenic black carbon and sulphate emissions. However, observations show no sign of a significant decreasing trend in VWS (Fig. 1b), in contrast to the simulated, aerosol-induced decreasing trend in ref. 1. We further show that the decrease of VWS in the recent epoch is caused by substantially advanced (by 15 days) tropical-cyclone occurrences, caused by the early onset of the Asian summer monsoon. There is a Reply to this Brief CommunicationArising by Evan, A. T. et al. Nature 489, doi:10.1038/nature11471 (2012). About 90% of the Arabian Sea pre-monsoon tropical cyclones occur from mid-May to mid-June, during which the mean VWS over the tropical-cyclone intensification zone increases approximately from 12m s to 25ms (Fig. 1c). We discover that the mean date of the lifetimemaximum intensity (LMI) of tropical cyclones occurred 15 days earlier from 8 June to 24 May (Fig. 1c) and the mean genesis date of tropical cyclones also shifted earlier by 16 days from 6 June to 21 May. The epochal changes of themean tropical-cyclone genesis andLMIdates are significant at 99.8% confidence level by Student’s t-test. Even if we exclude the two earliest storms in the recent epoch, the shift inmeanLMI (or genesis) date is 11 (or 13) days,which remains significant at above the 99% confidence level. Using the 15-days-earlier tropical-cyclone LMI date results in a remarkable reduction of the storm-ambient VWS, by about 5.8m s from before 1997 to the recent epoch (Fig. 1c), which is an order ofmagnitude larger than themodel-simulated, aerosol-induced VWS decrease (a difference of about 0.5m s between the two epochs). What has caused the earlier occurrences of tropical cyclones in the recent epoch? The genesis of tropical cyclones requires favourable environmental conditions and the likelihood of genesis can be estimated by the Genesis Potential Index (GPI). Computation of the GPI averaged over themaingenesis region (8u–20uN,55u–75uE) reveals that themaximumGPI occurred significantly earlier in the recent epoch (Fig. 1c), which explains why the recent tropical cyclones occurred earlier. Further analysis of the factors controlling the GPI change indicates that the shift of maximum GPI can be attributed mainly to the increased low-level cyclonic shear vorticity, which enhances the boundary-layermoisture convergence and lower tropospheric humidity. Meanwhile, the contribution of ambient VWS to GPI change is moderate. As shown in Fig. 1d, the monthly mean relative vorticity of 850 hPa inMay, duringwhich all tropical cycloneswere generated in the recent epoch, has significantly increased from before 1997 to the recent epoch. This finding is consistent with the early-onset trend in the Asian summer monsoon and with May (or June) storms being associated with an early (or late) onset of the southwesterly monsoon. We further find that the early development of the southwesterly monsoon may be caused by enhanced land–ocean thermal contrast between the Asian landmass and the equatorial Indian Ocean (Fig. 1d), which can reinforce the northward pressure gradients that in turn strengthen southwesterly monsoon and associated cyclonic shear vorticity. We consider that the increased land–ocean thermal contrast may be attributed to internally generated interdecadal variation and anthropogenic warming. On the one hand, the upward trends seen in tropical-cyclone intensity, low-level vorticity and the land–ocean thermal contrast all display a significant rapid upswing in the late 1990s (Fig. 1), and these upswings are in phase with the swift phase transition of the Interdecadal Pacific Oscillation, suggesting that natural variability could be the cause. On the other hand, previous studies have suggested that (1) increasing greenhouse-gas warming will enhance land–ocean thermal contrast in summer and (2) the