Dynamical Consequences of Polar Amplification on Standing Rossby Waves: a Laboratory Perspective

TL;DR

This study probes the dynamical consequences of polar amplification on mid-latitude Rossby-wave behavior using 24 independent experiments in a large rotating annulus, where Δ{T}, U_{b}, and β are independently controlled. The results show that large-scale standing Rossby-wave structures and their zonal wavelength follow the canonical relation $\\lambda = 2\\pi \\sqrt{U_{b}/\\beta}$ and are relatively insensitive to Δ{T}, while small-scale, transient activity is sensitive to Δ{T}$ through the buoyancy frequency $N$ and the energy spectra. A key finding is that the partitioning of variability into standing versus transient components collapses when expressed in terms of the non-dimensional group $U_{b}\\beta/N^{2}$, linking three governing parameters to the observed dynamics. The findings imply that polar amplification could promote more persistent mid-latitude weather but with decreasing zonal wavelength and north-south extent of persistent events, highlighting the nuanced balance between wind, stratification, and PV gradient in shaping future mid-latitude climate variability. These laboratory insights help disentangle coupled geophysical feedbacks and offer a framework for interpreting how polar amplification might modulate the frequency and character of extreme weather events.

Abstract

Polar amplification describes the predicted reduction in the latitudinal surface temperature gradient, which will have physical implications for mid-latitude dynamics. The precise nature of these dynamical consequences remains unclear. Here we explore aspects of polar amplification by way of 24 distinct idealised laboratory experiments. The apparatus employed can independently prescribe laboratory analogues for the latitudinal temperature gradient ($Δ{T}$, which controls the stratification $N$), background zonal flow speed ($U_{b}$), and strength of the background gradient in potential vorticity ($β$). The ability to control these processes individually is beneficial as decoupling them from one another enables their influences can be examined separately. Reducing the sidewall temperature difference substantially reduces small-scale and high frequency dynamics, but does not affect the large scale features of the flow, including the north-south amplitude of standing meanders. Reducing the zonal flow speed does reduces the length-scales and amplitudes of the standing Rossby waves, while reducing the potential vorticity gradient has the opposite effect; these responses are well described by the canonical expression relating the standing Rossby wavelength to $\sqrt{U_{b}/β}$. Variability is partitioned into components that are standing and transient; the response of this variability partitioning depends on all 3 experimental parameters, and a non-dimensional term is developed ($U_{b}β/N^{2}$) which captures the behaviour of the variability. These findings suggest that the dynamical consequences of polar amplification is a tendency for mid-latitude weather to shift away from transient storms towards more persistent events, however the zonal wavelength and north-south extent of these persistent events will tend to decrease.

Figures (11)

• Figure 1: Panel a) is a schematic of the Large Rotating Annulus (LRA). Our experiments vary the annulus rotation rate $\Omega_{a}$ (clockwise), the topography rotation rate $\Omega_{b}$ (anti-clockwise), and sidewall temperature difference $\Delta{T}$. Here, $\Omega_{a}$ varies the potential vorticity gradient $\beta$ according to Equation \ref{['eqn:beta']}, and $\Omega_{b}$ varies the zonal flow speed $U_{b}$ relative to the bump as $U_{b}=r_{b}\Omega_{b}$. Panel b) is a schematic of a radial transect across the LRA. The resting reference depth is represented by the horizontal dotted line at $D=$0.25 m; the parabolic free surfaces of the different $\beta$ cases are shown by the solid coloured lines. A cross-section of topographic bump is included for scale. Panels c) and d) show the parameter spaces for experiments that vary the zonal flow speed $U_{b}$ and potential vorticity gradient $\beta$, respectively. In both cases, 2 sidewall temperature difference values are used: $\Delta{T}>8^{\circ}$C and $\Delta{T}\approx2^{\circ}$C, referred to as large and small $\Delta{T}$, respectively. The background colours and black contours indicate the predicted zonal modenumber of the standing Rossby waves based on the respective $U_{b}$ and $\beta$ values.
• Figure 2: A sequence of images taken at regular time intervals after passive tracer dye is released into the equilibrated flows of experiments with $U_{b}=$25 mm s$^{-1}$ and $\beta=2.8$ rad m$^{-1}$ s$^{-1}$, for small and large sidewall temperature differences (top and bottom rows, respectively). Note that the large scale structures of the different cases are remarkable similar; however, the larger $\Delta{T}$ case exhibits substantially more small scale features and distribution of the passive tracer. These images are available in video form on the FluidsIn4K youtube channel at www.youtube.com/watch?v=CatIj6DU6ss
• Figure 3: The upper 2 rows show snapshots of surface temperatures for experiments that vary the zonal flow speed $U_{b}$ with small and large sidewall temperature differences $\Delta{T}$ (top and second row, respectively). The colourmap in each panel is centered about the snapshot mean of the annulus temperature and spans a range given by the snapshot standard deviation of the annulus temperature; these values are indicated in the lower left of each panel. The lower 2 rows show the time averaged surface temperature anomalies relative to the location of the bump for the same set of experiments as the upper 2 rows. The colourmaps are centered about zero and span a range given by the annulus standard deviation, which is indicated in the lower left of each panel.
• Figure 4: Time average surface temperatures of the experiments that vary $U_{b}$ for small and large $\Delta{T}$ (top and bottom rows, respectively); here the temperature fields have been transformed into radial--azimuthal coordinates. The annulus mean temperature of each case is contoured in black, and the approximate location of the bump is indicated by the white circle. The colourmap is centered about the annulus mean temperature and spans a range given by the standard deviation; these values are included in the upper right of each panel.
• Figure 5: The effect of varying the potential vorticity gradient $\beta$ on surface temperature snapshots (upper 2 rows) and standing temperature anomalies (lower 2 rows); these panels have the same layout and colourmap convention as per Figure \ref{['fig:03_SST_snaps_means_ub']}.