MRI: Acquisition of a Computing Cluster for Atmospheric and Geophysical Research

This is a Major Research Instrumentation (MRI) award which funds the acquisition of a 5-node cluster computer to support research and education activities in atmospheric and oceanic sciences. On the atmospheric side the primary application of the machine is the Extended Whole Atmosphere Community Climate Model (WACCM-X), which is used here to study the mechanisms through which sudden stratospheric warming events (SSWs) affect the ionosphere. SSWs are large-scale disturbances of the stratospheric circumpolar circulation in winter (typically over the North Pole) in which stratospheric temperature can rise by up to 50C over the course of a few days. While these disturbances occur over the pole and at altitudes of about 20km, they can affect the electron content of the ionosphere over the equator at an altitude of 300km, where they can interfere with satellites used for navigation, communication, and other purposes. The mechanisms by which this influence propagates over such great distances is not known but is believed to involve the vertical propagation of atmospheric tides. Work here tests the idea that the SSW impact can occur because westerly propagating planetary-scale waves excited by the SSW produce zonal wind anomalies which affect the vertical propagation of tides, and the alternative hypothesis that the mechanism involves changes in the heating due to ultraviolet absorption by ozone.

The oceanographic work considers wave-turbulence interactions in the coastal ocean, an important issue for understanding the transport of sediments associated with erosion and pollution and for understanding water properties essential to ocean biology. The work uses a large eddy simulation (LES) model, the Spectral Multi-domain model, to simulate waves in the near-shore ocean (bottom depths up to 100m or so). A key issue in understanding the interaction of waves and turbulence is the problem of separating the motion field into wave orbital motions and turbulence, which the PIs are attacking using a proper orthogonal decomposition (POD) in which individual modes represent different length scales within the flow and have energy levels which follow the classical Kolmogorov cascade. The cluster computer is used to calculate a three-dimensional POD which provides separate sets of modes to represent the wave orbital motions, small-scale turbulence, and nonlinear interactions between the two occurring at intermediate spatial scales.

The MRI award has broader impacts by providing infrastructure for research and education, including courses in oceanic, atmospheric, and computer science. In particular, the computer supports the newly formed PhD program in Coastal and Marine System Science, the first doctoral program at the university. Also, undergraduate students are strongly engaged in the assembly and maintenance of the cluster computer, thus acquiring hands-on skills in working with hardware integration and operating systems.