Abstract

Turbulent flows are characterised by chaotic behaviour spanning a wide range of length scales. For structures exposed to turbulent environmental loading, direct numerical simulation (DNS) of the Navier–Stokes equations is often infeasible due to prohibitive computational costs. An alternative approach relies on stochastic full-field simulators that reproduce key statistical features of turbulence through its spectrum. This talk presents our recent work on developing a novel turbulence spectrum model that captures the complex dynamics of long-range dependence and fractal characteristics prevalent in riverine and atmospheric boundary layer (ABL) flows, which are often overlooked in conventional models. I first introduce the parametric spectrum model for the streamwise velocity component, highlighting its asymptotic scaling behaviour and its connection to fractal and long-range dependent properties of the flow. Building on this, I present a theoretical approach that derives an intermediate −1 spectral scaling in the energy-containing regime, commonly observed near boundary layers, and examine its relation, or lack thereof, to Kolmogorov’s well-known −5/3 scaling in the inertial subrange. Finally, I discuss the interplay between the fractal and long-range dependence characteristics of the streamwise and lateral velocity components, outlining the implications for full vector-field simulation of turbulent flows.