LAUSR

Katabatic flows are notoriously difficult to model for a variety of reasons. Notably, the assumptions underpinning Monin‐Obukhov similarity theory (MOST) are inherently violated by the sloping terrain, causing the traditional flux‐gradient relations used in numerical weather prediction models to break down. Focusing on turbulent momentum transport, we show significant flux divergence, further violating MOST assumptions, and that the traditional parameterizations fail even with local scaling for katabatic flow. In response, we propose a modified local‐MOST stability‐correction function, informed by near‐surface turbulence observations collected over two mountainous slopes with inclination angles ( α ) of α≈7.8° and α≈35.5° . The proposed relation includes α directly, making data from both slopes collapse with unprecedented agreement. RMSE between measured fluxes and estimates from the proposed and Businger et al. (1971, https://doi.org/10.1175/1520-0469(1971)028<0181:FPRITA>2.0.CO;2) relations show significant improvement. Results can be used to inform future development of wall‐model and turbulence closures in the katabatic flow layer.