Implementation of a synthetic inflow turbulence generator in idealised WRF v3.6.1 large eddy simulations under neutral atmospheric conditions

Abstract

Abstract. A synthetic inflow turbulence generator was implemented in the idealised Weather Research and Forecasting large eddy simulation (WRF-LES v3.6.1) model under neutral atmospheric conditions. This method is based on an exponential correlation function and generates a series of two-dimensional slices of data which are correlated both in space and in time. These data satisfy a spectrum with a near “-5/3” inertial subrange, suggesting its excellent capability for high Reynolds number atmospheric flows. It is more computationally efficient than other synthetic turbulence generation approaches, such as three-dimensional digital filter methods. A WRF-LES simulation with periodic boundary conditions was conducted to provide prior mean profiles of first and second moments of turbulence for the synthetic turbulence generation method, and the results of the periodic case were also used to evaluate the inflow case. The inflow case generated similar turbulence structures to those of the periodic case after a short adjustment distance. The inflow case yielded a mean velocity profile and second-moment profiles that agreed well with those generated using periodic boundary conditions, after a short adjustment distance. For the range of the integral length scales of the inflow turbulence (±40 %), its effect on the mean velocity profiles is negligible, whereas its influence on the second-moment profiles is more visible, in particular for the smallest integral length scales, e.g. those with the friction velocity of less than 4 % error of the reference data at x/H=7. This implementation enables a WRF-LES simulation of a horizontally inhomogeneous case with non-repeated surface land-use patterns and can be extended so as to conduct a multi-scale seamless nesting simulation from a meso-scale domain with a kilometre-scale resolution down to LES domains with metre-scale resolutions.

Impact of Momentum Perturbation on Convective Boundary Layer Turbulence

“…In addition, we also perform one simulation with no inflow perturbations (referred to as No-SCPM) and another simulation with periodic boundary conditions (referred to as periodic) for context. By using periodic boundary conditions, it is implicitly assumed that both the atmospheric fields and the underlying land usage are encountered periodically during the simulation (Mirocha et al, 2014;Zhong et al, 2021). These simulations are summarized in Table 1 and described in more detail in Section 2.2.…”

A new wake detection methodology to capture wind turbine wakes using adaptive mesh refinement and Large Eddy Simulation

“…Different turbulent inlet boundary conditions for LES exist. The four most renowned methods are the Wind Tunnel Replication Method [24], the Recycling Method [25], the Precursor Database Method [26] and the Synthetic Turbulence Generator Method [27]. The wind tunnel replication method is the most straightforward method, easy to implement with high accuracy, although it is time-consuming and computationally expensive.…”

Learning the structure of wind: A data-driven nonlocal turbulence model for the atmospheric boundary layer

“…consists of the spectrum amplitude C := c 2 0 ε 2/3 /u 2 * , the characteristic length and time sales, L and T , respectively, the exponent ν and the weights θ NN of the neural network (20).…”

“…where N denotes the total number of weights of the neural network (20). The term accelerates convergence and also helps to avoid overfitting.…”

“…Such models are also important for coupling meso-and microscale models and providing inflow data for high-resolution large eddy simulations. [13][14][15][16][17][18][19][20] Turbulence has long been analyzed from a statistical point of view. 21 We seek to continue that enterprise with the assistance of new machine learning (ML) technology.…”