Sensitivity of the simulation of thermally-driven circulations in an idealized valley to planetary boundary layer parameterizations

Abstract

<p>This contribution aims at presenting results from the project “Atmospheric boundary-layer modeling over complex terrain”, a collaboration between the University of Trento, the University of Bolzano and the University of Innsbruck with the objective to evaluate the performance of turbulence and land surface parameterizations over mountainous terrain and to identify potential issues that have a large impact on model results and consequently on the quality of weather forecasts.</p><p>A set of Reynolds-averaged Navier-Stokes (RANS) simulations at 1 km horizontal resolution is performed in an idealized three-dimensional valley-plain topography, using typical geometrical features of a north-south Alpine valley, with ridges up to 1500 m above the valley floor and a distance of 20 km from crest to crest. Simulations are initialized with a linear and stable vertical profile of potential temperature, dry air and an atmosphere at rest. The aim of the modeling experiment is to evaluate the sensitivity of model results to planetary boundary layer (PBL) parameterizations, exploring the performance of the PBL schemes implemented in the Weather Research and Forecasting (WRF) model, including a newly developed k-ε closure. Results from the RANS simulations are compared against a large-eddy simulation (LES) with a resolution of 100 m, which is taken as the benchmark. A full diurnal cycle has been considered for the evaluation of numerical results, focusing on the development of along- and cross-valley thermally-driven circulations and on the associated thermal field both in the nighttime and in the daytime phases. The sensitivity of model results to the change of the PBL scheme is assessed using as key metrics the strength and the timing of the thermally-driven circulations, as well as the vertical profiles of mean and turbulent quantities, when available. Results show that in most cases there is a good agreement between RANS simulations and the LES considering the main features of both along- and cross-valley circulations and the diurnal evolution of the thermal stratification. In particular, the intensity of the along-valley wind is generally well-reproduced by all the RANS simulations, while higher discrepancies are found for the timing of the evening transition. On the other hand, RANS simulations are in good agreement with the LES considering the timing of slope winds, whereas the simulation of their intensity presents much more variability, especially during nighttime.</p>