Spatiotemporal Variability of the Global Ocean Internal Processes Inferred from Satellite Observations

Abstract

Using a new analysis tool, namely, multiscale window transform (MWT), and the MWT-based theory of canonical transfer, this study investigates the spatiotemporal variations of the nonlinear interactions among the mean flows, interannual variabilities, quasi-annual fluctuations, and eddies in the global ocean. It is found that the canonical kinetic energy (KE) transfers are highly inhomogeneous in space, maximized in the western boundary current (WBC), Southern Ocean, and equatorial regions. In contrast to the equatorial and WBC regions where the temporal KE cascades are mainly forward, the Southern Ocean is the very place where coherent large-scale patterns of inverse KE cascade take place. The canonical transfers are also found to be highly variable in time. Specifically, in the Kuroshio Extension, the transfer from the mean flow to the interannual variability is in pace with the external winds from the eastern North Pacific; in the subtropical gyre, the mean flow-to-eddy transfer is responsible for the variability of the eddy kinetic energies (EKE) at both interannual and seasonal scales; in the tropics, the downscale transfers to the eddies from the other three scales all contribute to the interannual modulation of the EKE, and these transfers tend to decrease (increase) during El Niño (La Niña) events. In the Southern Ocean, the high-frequency eddies are found to feed KE to the low-frequency variability through temporal inverse cascade processes, which have been strengthened due to the enhanced eddy activities in the recent decade. Also discussed here is the relation between the seasonal EKE variability and the eddy–quasi-annual fluctuation interaction.

Overestimated Eddy Kinetic Energy in the Eddy‐Rich Regions Simulated by Eddy‐Resolving Global Ocean–Sea Ice Models

“…For instance, Feng et al (2017) present that a more energetic mesoscale eddy activity is simulated in the South China Sea, possibly associated with the overestimated Kuroshio intrusion (Feng et al, 2020). Generally speaking, eddy-resolving models are prone to simulate stronger eddies in regions with strong currents, such as WBCs in the Northern Hemisphere (Fu & Smith, 1997;Rieck et al, 2015;Sasaki et al, 2020;Williams et al, 2011;Yang & San Liang, 2019;Yang et al, 2017), eastern tropical oceans (Chao & Fu, 1995), and in the Southern Ocean (Dietze et al, 2020;Hristova et al, 2014;McClean et al, 2011). Therefore, the simulated mesoscale eddies in eddy-resolving OGCMs are not uniformly weaker than observations and worth to be investigated further at the regional scale.…”

On the Errors of Estimating Oceanic Eddy Kinetic Energy

“…The need to study the energetics in terms of time‐mean is also relaxed thanks to the instantaneous orthogonality. One good example of such technique is the recently developed Multiscale Window Transform and the related Localized Multiscale Energy Analysis (Liang, 2012, 2016; Liang & Anderson, 2007; Liang & Robinson, 2005, 2007; Yang & Liang, 2019), which has already been used in eddy and mean‐flow energy analysis in the Kuroshio and Kuroshio Extension regions (Y. Yang et al., 2017; Yang & Liang, 2016, 2018). The orthogonality of such decomposition methods ensures exact examination of energy transfer between scales, an advantage over traditional methods using either spectral truncation or coarse‐graining (Aluie et al., 2018; Arbic et al., 2013; Scott & Wang, 2005; Tulloch et al., 2011).…”

“…The need to study the energetics in terms of time-mean is also relaxed thanks to the instantaneous orthogonality. One good example of such technique is the recently developed Multiscale Window Transform and the related Localized Multiscale Energy Analysis (Liang, 2012(Liang, , 2016Liang & Anderson, 2007;Liang & Robinson, 2005Yang & Liang, 2019), which has already been used in eddy and mean-flow energy analysis in the Kuroshio and Kuroshio Extension regions (Y. Yang et al, 2017;Yang & Liang, 2016, 2018.…”

Time‐Dependent Role of Multiyear La Niña in Impacting the Pacific Tropical Instability Waves

“…This practice is problematic because $${\left[\overline{u}(t)\right]}^{2}+{\left[{u}^{\prime }(t)\right]}^{2}$$, that is, the total energy is not conserved. A simple example of the multiscale energy representation problem is shown in the appendix of Yang and Liang (2019). In fact, multiscale energy is a concept in phase space, such as that in the Fourier power spectrum.…”

“…During the developing year of La Niña events, the activity of TIWs is more intense due to the strengthened equatorial cold tongue, enhancing the meridional temperature gradient and thus making the background flow more baroclinically unstable. The opposite is true during El Niño events (An, 2008; Contreras, 2002; Yang & Liang, 2019; J.‐Y. Yu & Liu, 2003).…”