Turbulent exchanges between the atmosphere and the underlying surface transport heat, moisture, momentum and pollutants, and thus understanding them is crucial to water resource management, meteorological predictions, climate modeling, wind energy production and pollution transport modeling and mitigation. The majority of established theories and techniques used to model, predict and observe turbulent exchange assume the underlying surface is flat, uniform and homogeneous. Consequently, in mountainous regions, turbulent exchanges are difficult to model and predict because they are intrinsically linked to the underlying surface conditions and complexity. Few observational studies have focused on the vertical structure of turbulent fluxes over mountainous terrain due to practical challenges. In sum, researchers and practitioners urgently need better theories and techniques to study and model turbulent exchange in alpine environments. In this dissertation, we analyze turbulent exchange measured over a steep, alpine slope in Switzerland to make quantitative comparisons with standard flat-terrain âlawsâ and practices and develop new theories and techniques adapted to mountainous terrain. First, flux-profile relations derived from Monin-Obukhov similarity theory are widely used in a variety of models to relate the surface conditions to atmospheric variables. We show that for katabatic flow, or thermally-driven air drainage that often occurs at night over sloping terrain, these relations break down because they do not account for the strong vertical gradients in the observed turbulent fluxes. However, the measurements exhibit clear functionalities that we use to derive new, empirical flux-profile relations for katabatic flow. Our functions indicate increased turbulent mixing compared to the flat-terrain relations. Misalignment between gravity and the inclined surface helps explain these functional differences because it reduces (and even reverses over steep slopes) the typical nighttime buoyant suppression of turbulent kinetic energy, which acts vertically via gravity. We also revise the governing flow equations to properly account for this misalignment. Second, standard data treatment techniques, such as sensor tilt corrections, do not hold true over complex topography. We develop an optimized sector-wise planar fit (SPF) tilt correction to best select the SPF input parameters for any complex site, and quantify the resulting sensitivities in the momentum fluxes. Finally, we present a mechanistic study that combines measurements and the one-dimensional momentum balance. Results show that frictional and buoyant mechanisms dominate, but that drag forces from the alpine grass canopy and the modeled outer-layer pressure (for conditions with low wind speeds) are also significant. In summary, this one-dimensional model or the newly developed flux-profile relations could replace inappropriate, flat-terrain wall models in larger-scale numerical models to improve atmospheric predictions for katabatic flow regimes. Additionally, terrain appropriate sensor tilt corrections and proper alignment of buoyant forces improve turbulent flux estimates for complex and steep topography, and hence have ramifications for similarity or budget analyses that require accurate quantification of turbulent exchange.