PhD Final Defense – Hope Hunter

Spatial Structure as a Source of Information in Arctic Atmospheric Research

Advisor: Hannah Horowitz

Location: Zoom

https://illinois.zoom.us/j/85167939709?pwd=KpMxrxBVAax3fbVV5a8vxcbnAW9mtY.1

Meeting ID: 851 6793 9709 Passcode: 683726

Abstract

The Arctic exerts a large influence on global climate, yet it remains among the most sparsely observed regions of the atmosphere. Where observations are this limited, appropriate methodology is essential to extract insight from them. However, the spatial structure of Arctic data is too often overlooked in analysis despite its value as a source of information. This dissertation examines how the explicit representation of spatial relationships, spatial organization, and geometric structure can reveal features of Arctic atmospheric processes that conventional diagnostics obscure. The argument is developed across four studies spanning two regions of the Arctic atmosphere: the air–sea interface and the stratospheric polar vortex.

The first two studies regard surface energy exchange in a summertime Arctic environment. Using in situ measurements from autonomous saildrone platforms, the first evaluates operational deterministic and ensemble forecasts of turbulent sensible and latent heat fluxes and attributes flux errors to errors in the underlying atmospheric and oceanic state variables, observing spatial organization in errors. The second study reframes forecast error as a spatially organized quantity rather than a set of independent residuals, developing a spatial-statistical framework for model validation across misaligned spatial scales that separates structural, global, and local sources of error.

The final two studies concern the three-dimensional structure of the stratospheric polar vortex. The third study describes a topological and geometric algorithm which identifies the polar night jet wind and vortex body by geopotential edge, and produces a compact set of morphology metrics that zonal-mean indices cannot capture. These metrics drive a state-machine classifier that labels vortex states across each winter season and detects and characterizes sudden stratospheric warmings by morphology and severity. The output of this algorithm is utilized in the fourth study to evaluate the structure of the stratospheric polar vortex across 45 seasons, finding insights into its climatological morphology and interannual variability. This uncovers how the stratospheric polar vortex evolves throughout the winter and quantifies the true frequency of sudden stratospheric warmings in the Arctic.

Together, these studies contribute new knowledge of Arctic atmospheric processes alongside transferable, spatially explicit methods: a spatial-statistical framework for model validation across misaligned spatial scales, and a morphological algorithm for the characterization and classification of the polar vortex.