Turbulent heat fluxes and katabatic flow at glacier surfaces : insights from measurements and modelling

Turbulent heat fluxes are critical components of glacier surface energy balance models but remain challenging to accurately simulate, particularly during shallow katabatic flow – buoyancy-driven downslope winds prevalent over glaciers in summer. These challenges arise because bulk aerodynamic methods and existing one-dimensional katabatic wind models oversimplify turbulent mixing, and consequently do not accurately simulate turbulent heat fluxes when wind speed maxima occur near the surface. This thesis addresses these limitations by combining multi-year field observations from a glacier in Yukon, Canada, advanced data processing techniques, and numerical modelling to improve the characterization of katabatic flows and turbulent heat fluxes at glacier surfaces. First, multi-height eddy covariance and meteorological data were analyzed to assess bulk aerodynamic methods under different near-surface flow regimes. A novel processing method for single-level eddy covariance data was developed, ensuring the validity of flux calculations during variable flow conditions and low-level wind speed maxima. This method significantly improved agreement between modelled and observed fluxes, particularly in shallow katabatic flow regimes. Second, cospectral analysis of the eddy covariance data revealed a strong correlation between the height of the katabatic jet – confirmed through atmospheric profiling with tethered kite measurements – and peaks in streamwise heat flux cospectra. Turbulent mixing lengths for momentum and heat deviated from standard “law-of-the-wall” scaling, reflecting strong near-surface stability and a sharp thermocline coinciding with the wind speed maximum. These findings demonstrate that the katabatic boundary layer differs substantially from canonical slope flow models. Finally, a two-dimensional katabatic wind model was developed to simulate the suppression of turbulence due to near-surface stability and the localized entrainment of warmer air across the thermocline, which dominates heat transfer processes. Numerical simulations and asymptotic analysis revealed key characteristics of katabatic flow, including velocity and temperature profiles, entrainment rates, and turbulent heat fluxes. By integrating field measurements, data analysis, and modelling, this thesis enhances the process- based understanding of katabatic flow and turbulent heat fluxes, improving surface energy balance models and advancing the accuracy of glacier melt simulations.