Unique Windward Measurements and a Mesoscale Simulation of an Extremely Long-Lasting Severe Bora Event

Abstract

Unique data from a 100-m meteorological mast located on the windward side of the Dinaric Alps, Croatia, are compared to high-resolution Weather Research and Forecasting (WRF) model simulations. This was performed for an especially strong and long-lasting (more than 20 days) wintertime bora event. The agreement between the measurements on the mast and the respective WRF simulation was generally very good, even with respect to the time series of the turbulence kinetic energy. Based on this finding, which validates the WRF model suitability for numerical simulations of transient winds in windward areas, this approach can be used in future studies to explore the severe bora upwind of the coastal mountains, which has been studied inadequately thus far. In this context, some of the preliminary results are outlined here.

Appendix

For each selected cross-section, the atmospheric inversion height h(x) was determined visually from the plots of the potential temperature θ as the height of isoline above the region of highest θ gradient. After this curve was determined, it was normalized by h₀, its average value in the first 15 km of the cross-section on the windward side of the flow, where the flow had yet to accelerate. The layer-mean wind speed u(x) was normalized by u₀, defined in the same manner as h₀. The Froude number is defined as

$$ Fr = \frac{u}{{\sqrt {g^{\prime}H} }} = \frac{u}{{\sqrt {g\left( {\Delta \theta /\theta_{0} } \right)H} }} , $$

where g′ is the reduced gravity, H(x) is the bora flow thickness, θ₀ is the average value of θ in a layer that has its top at h(x) and has thickness of 0.4 H (rough yet robust estimate of inversion layer thickness for the chosen cross-sections) and Δθ is the total jump in θ over that layer. Thus, for each cross-section we obtain a curve for each of the u/u₀, h/h₀ and Fr variables.