Generation of westerly wind bursts by forcing outside the tropics

Sci Rep. 2021 Jan 13;11(1):912. doi: 10.1038/s41598-020-79655-7.

Abstract

The westerly wind burst (WWB) is an important triggering mechanism of El Niño and typically occurs in the western Pacific Ocean. The Fourier spectrum of the wind field over the western tropical Pacific is characterised by a large variety of peaks distributed from intra-seasonal to decadal time scales, suggesting that WWBs could be a result of nonlinear interactions on these time scales. Using a combination of observations and simulations with 15 coupled models from the Coupled Model Intercomparison Project Phase 6 (CMIP6), we demonstrate that the main drivers initiating WWBs are quantifiable physical processes rather than atmospheric stochastic signals. In this study, ensemble empirical mode decomposition (EEMD) from the Holo-Hilbert spectral analysis (HHSA) is used to decompose daily zonal winds over the western equatorial Pacific into seasonal, interannual and decadal components. The seasonal element, with prominent spectral peaks of less than 12 months, is not ENSO related, and we find it to be strongly associated with the East Asian monsoon (EAM) and cross-equatorial flow (CEF) over the Australian monsoon region. The CEF is directly related to the intensity of the Australian subtropical ridge (STR-I). Both the EAM and CEF are essential sources of these high-frequency winds over the western Pacific. In contrast, the interannual wind component is closely related to El Niño occurrences and usually peaks approximately two months prior to a typical El Niño event. Finally, the decadal element merely represents a long-term trend and thus has little to no relation to El Niño. We identified EAM- and CEF-induced westerly wind anomalies in December-January-February (DJF) and September-October-November (SON). However, these anomalies fade in March-April-May (MAM), potentially undermining the usual absence of WWBs in the boreal spring. Similar results are found in CMIP6 historical scenario data.

Publication types

  • Research Support, Non-U.S. Gov't