The Response of the Mid-Latitudes to Idealized Orography in the Presence of a Jet

Orography exerts a strong control on the zonally asymmetric component of atmospheric circulation through its effects on planetary-scale stationary waves. Understanding how orography does this is a complex problem, as the stationary waves which the orography excites can interact with other atmospheric phenomena, such as independently excited stationary waves, transients and critical layers, as well as nonlinearly with themselves. Here we revisit this problem by forcing the GFDL spectral dynamical core with idealized orography consisting of Gaussian "mountains". These mountains are located near the center of the midlatitude jet and have a wide range of maximum heights, from 200m to 4000m. In contrast to previous studies, we see distinct modes of response. The smallest mountains produce a highly zonal, wavenumber five response, resembling the "circumglobal" wave which has been seen in other contexts but has not been previously observed in studies of orographic forcing. As the height of the mountain is increased the circumglobal wave begins to resonate with itself and the amplitude of the response increases rapidly before saturating as the wave starts to break. At the same time, a second wave appears; this wave is longer than the circumglobal wave and has a different vertical structure. It propagates both zonally and meridionally, and resembles the response typically seen in other studies of orographic forcing. There is no evidence of the two waves interacting with each other.

To help interpret this behavior, we repeat our experiments in a barotropic model on a sphere with an idealized zonal-mean zonal wind profile similar to that of the dynamical core. In this setting an analogous transition is observed from a response dominated by the circumglobal wave to one dominated by the meridional wave. The simplicity of the model allows us to extensively explore this transition, which appears to take place in four stages. In the first stage the response of the wave is linear in the height of the forcing. It is unclear if our dynamical core experiments are ever in this regime. Next, the circumglobal wave begins to resonate with itself as the height of the mountain is further increased and so the amplitude of the response increases rapidly. However, conservation of wave activity implies that the amplitude of the response must saturate, which happens in the third stage. During this stage the meridional wave increasingly dominates the response. Finally the circumglobal wave is no longer excited and only the meridional wave is present; this stage was not reached in our dynamical core experiments.

These responses are sensitive to the width and strength of the jet, which in term determine the jet's "carrying capacity". For instance, simulations with broader jets do not produce a circumglobal response. As such, we use the barotropic model to map out the parameter space and develop scaling arguments to predict the heights at which the transition occurs. A limitation of our study is that, due to the idealized setting, we have considered the waves in a zonally symmetric base state, without the complications of other zonal asymmetries with which they could interact. On the other hand, our results are likely to be applicable to other zonally asymmetric forcings, such as localized heating.