Evaluation of turbulent surface flux parameterizations for the stable surface layer over Halley, Antarctica

Calculated surface fluxes from seven surface layer parameterizations are verified against 45 months of observations from Halley, Antarctica, with a temporal resolution of 1 h. The surface layer parameterizations are taken from widely used numerical models including the National Center for Atmospheric Research (NCAR) Community Climate models CCM2 and CCM3, the U.K. Met. Office Unified Climate Model, and the fifthgeneration Pennsylvania State University–NCAR Mesoscale Model (MM5). The observations include measurements of the mean wind speed and temperature inversion strength and direct measurements of the turbulent fluxes of heat and momentum. A comparison of the calculated and observed fluxes is conducted for conditions in which the surface layer is stably stratified. Based on these comparisons it is found that the simulated friction velocity values are adequate (although slightly larger than the observed turbulent fluxes) under all but the highest bulk Richardson number conditions (greatest static stability). In contrast the magnitude of the calculated sensible heat flux is frequently less than that of the observed sensible heat flux. The use of a larger scalar roughness length for heat compared to that for momentum is found to remove this bias in the calculated sensible heat fluxes. The correlation between the observed and calculated fluxes of heat and momentum is acceptable for the lower bulk Richardson number regimes, but is near zero for the high bulk Richardson number regime. The correlation between the calculated and observed fluxes is in general better for the momentum flux than for the sensible heat flux. The bias in the calculated sensible heat flux could have significant implications for numerical simulations in which the flow is driven by surface processes, and may pose problems for climate-scale simulations. The impact that errors of the observed magnitude have on simulated katabatic winds is explored with a series of twodimensional numerical simulations using MM5. Inferences about the relevance of these findings for climate simulations are also addressed

Details