An Aerodynamic Roughness Length Parameterization for Numerical Modeling of the Coastal Atmospheric Surface Layer

Parameterization of atmospheric surface layer (ASL) momentum, heat, and moisture fluxes is a critical component of any geophysical fluid numerical model. Virtually all numerical weather prediction (NWP) models parameterize ASL turbulence using the bulk-flux approach, which rely on empirical similarity relationships derived from observations of vertical gradients of winds, scalars, and turbulence over homogeneous surfaces. Recently, Hlywiak et al (2023) showed that NWP prediction over two months in a coastal region (Monterey Bay, CA) using the United States Navy's Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS^®*) was degraded, primarily because bulk-flux parameterizations are not designed for highly heterogeneous ASL environments such as the coastal and littoral zones. Specifically, modeled surface drag coefficients are independent of wind direction, in accordance with traditional one-dimensional ASL similarity theory. In contrast, observations revealed a distinct bifurcation in inland near-surface momentum fluxes depending on whether the flux footprint of the flow reflected oceanic or land origins. Here, we discuss an ongoing effort to improve coastal near-surface NWP by parameterizing the effects of three-dimensional processes – namely, advection and extinction of turbulent kinetic energy (TKE) – on sub-grid-scale turbulent ASL fluxes. Specifically, we consider the physics relevant to internal boundary layer formation and growth – which is ubiquitous within coastal regions – to develop a parameterization for the effective aerodynamic roughness length. We first reveal roughness lengths derived from turbulent flux measurements collected by an array of cross-coastal flux towers collected during several field campaigns as part of the Office of Naval Research-sponsored Coastal Land-Air-Sea Interaction (CLASI) Departmental Research Initiative. Results show that the derived aerodynamic roughness lengths – and consequently drag coefficients – within roughly six kilometers inland of the coast are controlled by whether the flow at the coast is onshore versus offshore. Results are less sensitive to ASL stability within the study region. We then derive an empirical relationship between the observed effective roughness lengths and key parameters identified from the observations – i.e., a downstream residence timescale and a TKE dissipation timescale – using dimensional analyes. We show that the empirical relationship can be fitted to standard equations for internal boundary layer heights developed by previous literature. Finally, large eddy simulation (LES) output will be used to validate the derived relationship, and the results will be discussed in the context of NWP modeling. This work addresses a key limitation in NWP modeling, which is the universal approach towards parameterization of ASL turbulent fluxes across both homogeneous and heterogeneous environments.

*COAMPS^® is a registered trademark of the U.S. Naval Research Laboratory