Validation of Multi-Scale Simulations of the Flow over Big Southern Butte Using Weather Research and Forecasting Model

Abstract:

Advances in high performance computational resources and frameworks now make possible the use of Numerical Weather Predication (NWP) models for high-resolution simulations of atmospheric flows. In order to develop best practices, standards, and procedures for multi-scale simulations, we need to carry out extensive validation of NWP models across unprecedented range of scales from hundreds of kilometers to tens of meters. However, there are limited observational data available for evaluating high-resolution models. Recently, Nunalee et al (2015) validated large-eddy simulations (LES) using WRF for flow and dispersion based on the Cinder Cone Butte experiment carried out in Idaho in 1982. This study involved moderately complex terrain. We now extend the study to a significantly more complex terrain based on a more recent field study in Idaho. This field study include two experiments: the first one carried out in 2010 and centered on the Big Southern Butte (BSB) and the second in 2011 centered on the Salmon River Canyon both in Idaho (Butler et al., 2015).

As a first step, here we focus on using the observations from the BSB experiment to validate multi-scale simulations using the WRF model. We carry out both mesoscale simulations and large-eddy simulations (LES). Nested mesoscale simulations are carried out using the innermost nest with grid cell size of 300m while nested WRF-LES are carried with grid cell size of ~50m. We analyze the performance of PBL scheme in mesoscale simulations and the resulting interplay between subgrid parameterization and numerical advection scheme in LES. The results of this analysis are used to assess performance of PBL schemes in complex terrain where the assumption of horizontal homogeneity on which these schemes are based are violated and to suggest the modifications to PBL scheme to account for the effect of heterogeneity.