A Three-Dimensional Portrait of the Multiscale Interaction Underlying the Typhoons over Northwest Pacific Ocean

The multiscale interaction analysis of a typhoon calls for an extension of the Lorenz cycle analysis to a highly localized, and moving framework. This, however, proves to be a very challenging problem, at least in two aspects. The first is from the non-stationary processes as a typhoon develops, which fails the existing Reynolds decomposition-based energetics methods. The second regards a faithful extraction of the internal energy transfer among scales from the intertwined nonlinear terms, which Plumb (1983) has argued long before that the current formalisms do not yield a unique separation, and hence the so-obtained physics is unclear and not convincing. This These issues have not been resolved until the advent of multiscale window transform (MWT) by Liang and Anderson (2007). Based on MWT, Liang (2016) developed a theory of canonical transfer, and a methodology called localized multiscale energy and vorticity analysis (MS-EVA), which have these issues resolved in a uniform treatment. In this study, these theories/methodologies are used to unveil the three-dimensional local multiscale dynamical processes underlying the composite typhoon over the northwest Pacific Ocean, in the hope of gaining more insight into the complex typhoon dynamics.

A total of 201 typhoon events occurring over the northwest Pacific Ocean region from 1995 to 2016 are analyzed. To have a general understanding of their dynamics, a composite of these events is constructed according to their locations and strengths as derived from the dataset of best track and strength from the Shanghai Typhoon Institute of China Meteorological Administration. The multiscale energetics are computed based on the ERA-5 dataset from European Center for Medium-Range Weather Forecasts. With MWT the variables are firstly reconstructed onto two scale windows, i.e., the basic flow (longer than 32 days) and typhoon window (shorter than 32 days). The pressure, winds, and temperature fields of the reconstructed typhoons on the typhoon window are consistent with that of the typical typhoons in previous studies. The energetics for the reconstructed typhoons are then calculated and composited.

By strength tendency, the lifecycle of the composited typhoon is divided into three stages, i.e., the development stage, maintenance stage, and decay stage. We found that the first two stages have almost the same energetics structures, except for magnitude. In the following we hence only give a description for the development and decay stages.

In the development stage, the main source of typhoon energy is the diabatic heating in the upper troposphere. As a result, the local available potential energy (APE) is increased and tends to be converted to kinetic energy (KE). The gained KE is then transported, via pressure work, from the upper troposphere to the lower troposphere, to make up the KE loss due to surface dissipation. In contrast, the vertical component of the KE transport process brings KE from the lower troposphere to the upper troposphere, but the magnitude is much smaller in comparison to its pressure work counterpart. Besides, the baroclinic instability of the upper tropospheric basic flow is also an energy source fueling the typhoon, by transferring APE from the basic flow to the typhoon. As a result, baroclinic processes are not negligible when typhoon dynamics is investigated. The above processes are almost axisymmetric, accounting for the axisymmetric dynamics. In the axis-asymmetric KE energetics balance, the canonical KE transfer dominates. As rigorously proved in Liang and Robinson (2007), this characterizes the perturbation energy growth in a barotropic flow, and hence is a natural measure of (nonlinear) barotropic instability. Just as in our previous case study (e.g., Wang and Liang, 2017), here it is also found to have a dipolar structure throughout the troposphere, with a positive center and negative center in its southeast and northwest part, respectively. In other words, it has a wavenumber-1 structure and hence accounts for the corresponding axis-asymmetric typhoon dynamics. Other processes also possess significant axis-asymmetric patterns, including the horizontal components of the KE transport and pressure work, which vary with heights. In the lower troposphere, KE is transported from the southeastern outer region (of the composite typhoon) to its eye region. In the mid-troposphere, the KE transport also has a wavenumber-1 structure, but here KE is transported from the southern half to the northern half, whereas in the upper troposphere, KE is transported from the southeast to the northeast. For pressure work, the corresponding patterns are almost opposite to their KE transport counterparts, implying that these processes are inter-correlated.

In the decay stage, the upper tropospheric energy sources as identified in the development stage no longer align with the typhoon; rather, they shift to its northeast by a significant distance. The structure which once allows the typhoon to efficiently extract energy no long exists. As a result, the typhoon tends to decay. The moral of this study is that, from an energetics viewpoint, whether the typhoon develops or decays is sensitive to the location of the energy source relative to its body. If they are aligned vertically, the typhoon can efficiently absorb the energy to fuel itself, leading to development; if the energy source deviates away from its body, the typhoon tends to decay. These observations demonstrate the necessity of extending Lorenz’ bulk energetics analysis to a localized framework; otherwise, the discrepancy between the locations of the energy sources and the typhoon will be disguised due to the spatial averaging.

Keywords: typhoon, multiscale window transform, localized Lorenz cycle, barotropic instability, baroclinic instability, canonical transfer, buoyancy conversion,