CGAFD Seminar "Exploring Mechanisms for Model-dependency of the Stratospheric Response to Arctic Warming" and "Exploring Physics-Dynamics Coupling With Moist Shallow Water Equations"

Speaker: Regan Mudhar, University of Exeter  

Title: Exploring Mechanisms for Model-dependency of the Stratospheric Response to Arctic Warming

Abstract: Recent studies propose that Arctic sea ice loss and associated warming influence wave propagation into the stratosphere, affecting the winter polar vortex. Through stratosphere-troposphere coupling, this may perturb the winter jet stream and affect surface weather. But the “stratospheric pathway” linking Arctic variability to midlatitude weather extremes is not well understood. State-of-the-art models, for example, do not show a robust stratospheric response to Arctic sea ice loss, in strength nor sign. We have used a dry idealised atmospheric modelling framework to better understand mechanisms and uncertainties in the stratospheric response to polar warming. We use Newtonian relaxation of temperature to a specified equilibrium temperature to simulate northern hemisphere winter, then alternatively force the model with an adjustable polar heating and modify the stratospheric state by adjusting polar vortex strength. Consistent with previous work, imposed polar warming increases upward wave propagation into the polar stratosphere, but with a notable dependency of the stratospheric response on heating depth and strength, as well as the vortex basic state. I will also show analysis of stratospheric responses in the more complex PAMIP (Polar Amplification Model Intercomparison Project) model simulations, proposing that our results from both the idealised and more complex models help to explain some of the model-dependency driving the range of simulated responses.  

Speaker: Nell Hartney, University of Exeter  

Title: Exploring Physics-Dynamics Coupling With Moist Shallow Water Equations

Abstract: At the heart of any weather or climate model is the dynamical core. This is the part of the model that computes solutions to the fluid equations describing the behaviour of the atmosphere, known as the dynamics. Any processes that have an important impact on the overall flow but are not captured by the dynamical core are known as the physics. The physics includes sub-grid-scale processes such as convection and non-fluid dynamical processes like radiation. Physics processes are typically parametrised and dealt with separately from the dynamical core, and then somehow coupled to the dynamics so that they can influence the solution correctly. The implementation of this so-called physics-dynamics coupling can have a significant impact on the effectiveness of a model and there are many challenges associated with it, including how the physics should be dealt with in the model’s timestep.

An ideal context to look at some of these physics-dynamics coupling questions is in a moist shallow water model. The traditional shallow water equations retain many pertinent features of a dry atmosphere but are simple and computationally cheap. Incorporating moisture into the shallow water equations introduces physics into what is otherwise a dynamics-only model, giving an inexpensive framework that couples a physics process to simple (but reasonably realistic) dynamics.

This presentation will describe our implementation of a flexible, unifying moist shallow water model which encompasses a number of approaches to moist shallow water modelling. I will also discuss our newly-developed moist shallow water model which incorporates all moist processes in the dynamics. This gives us a ‘ground truth’ - a model with no physics-dynamics coupling with which to compare coupling strategies against. Our models are built in the dynamical core toolkit Gusto, which follows a compatible finite element approach like that of the next generation Met Office model. I will describe how we have used moist shallow water test cases in Gusto to explore some of the implications of different moist shallow water modelling choices, as well as investigating questions about physics-dynamics coupling and how physics is dealt with by a model’s time-stepping scheme.