A Turbulence Library for Asynchronous Coupling of Meso- and Microscale Models

To realistically model a wind farm, we need to extend the range of spatial and temporal scales represented in a model from 10’s of meters to 100’s of km’s and time scales of few seconds to days. These scales in the atmosphere are represented by either mesoscale or microscale models, which have different characterizations of various dynamical and physical processes. Setting time varying turbulent inflow boundary and initial conditions for a microscale simulation that is designed to produce an internal boundary layer is a key challenge. Improper inflow turbulent conditions could lead to transient conditions near the inlet (or inflow) boundary that may persist several boundary layer depths into the model domain. The inflow conditions that are necessary for stimulating the LES simulation of wind fields over a region include the mean velocity profile and a measure of turbulent kinetic energy.

We present a framework for building a library of precursor simulations for representative boundary conditions and then using the library to couple the evolution of atmosphere across WRF-ARW (mesoscale) to WRF-LES (microscale) models. The challenge here is the generation of a library of simulations that could be useful for most frequent meteorological conditions that would influence the boundary layer development at the microscale. As an initial test of the approach, we will mostly focus on regions away from coast and terrain changes that could have a significant influence on the internal boundary layer development. Ensemble of simulations are performed and evaluated for selected flow conditions at the SWIFT site in Northeastern Texas. Our approach addresses two potential sensitivity cases (a) atmospheric stability states, including neutral and convective boundary layers, and (b) issues concerning with the time and spatial resolution of the precursor runs. For each stability case, the three key variables are the mean wind fields, surface heterogeneity and surface heat fluxes. The second key issue for implementing the library relates to the spatial resolution of the LES model and the frequency at which we need to save the model outputs for building the library. We will present an initial construction of the turbulence library and results from test simulations using the asynchronous model coupling concept. We will also discuss the computational performance of the coupling strategy.