simple site

Urban LES intercomparison

Overview

Motivation for this project can be summed up in few key points

  • Bright future: microscale turbulence resolving models – essential element of urban services and planning
  • Aiding urban-canopy parameterizations in meso-scale & global-scale atmospheric models
  • How good is LES in reproducing urban boundary layer?
  • What approaches (numerics + physics) work best?

Participants

Currently there are 2 paticipant models:

  • LES from RCC LMSU

    Mixed dynamic subgrid closure + nesting Double-moment cloud microphysics & radiation (RRTM) modules Land surface coupling Atmospheric chemistry & aerosol transport CPU/GPU support via C/C++/MPI/OpenMP/CUDA

  • Fluidity from Institute of Atmospheric Physics CAS

    Basic Information for Fluidity: Numerics: Spatial discretisations: first-order and second-order Continuous Galerkin, Discontinuous Galerkin, Control Volume Temporal discretisation: first-order explicit, second-order semi-implicit and implicit Turbulence Package: Reynolds-Averaged Navier-Stokes (RANS) Modelling Generic length scale turbulence parameterization (based on two equations, for the transport of turbulent kinetic energy (TKE) and a generic second quantity) Standard k – ε Turbulence Model Large-Eddy Simulation (LES): Second-order dissipation model Fourth-order dissipation model Dynamic Smagorinsky model

to become a participant send an email to Andrey Debolskiy or Evgeny Mortikov

Stages

Idealized urban configurations

The preliminary setup follows JAS paper 1 here is a scheme for the experiments image

for testing non-stratified flow is considered. Here has detailed description of the idealized setup. Below is just a breif overview

  • Domain: length = 256 m, width = 128 m, height = 64.0 m;
  • Grid: nx = 512; ny = 256; nz = 128, so that uniform resolution of 0.5m in any direction is achieved
  • Flow configuration: Open Channel flow with constant external pressure gradient of dpdx = - phys.rho_ref * ustar_r^2 /(height - buildings_height), where phys.rho_ref = 1.25 kg/m^3 is reference air density, ustar_roof = 0.25 is target dynamic velocity just above the building roofs .
  • Other fluid characteristics: 
          f = 0.0;                                # coriolis frequency [1/s]

      nu = 1.25 * 0.00001;                    # kinematic viscosity [m^2/s]
      xi = (1.0 / 0.7) * nu;                  # thermal diffusivity [m^2/s]

      rho_ref = 1.25;                         # reference density of air [kg/m^3]

      g = 9.81;                               # gravitational acceleration [m/s^2]
      Theta_ref = 283.15;                     # reference temperature [K]

      # --- no buoyancy
      beta = 0.0;                             # = g * thermal expansion coefficient = g / Theta_ref [m/(K*s^2)]
  • Integration model time: 2hours
  • Buildings setup: surface dynamic and thermal roughness parameters z0m = 0.01 m, z0h = z0m/10.0. building_height = 16 m.
  • Building configuration: EXP1 
    h = 16 # [m], building height
      patch_1 {
              type = "box";

              xmin = 0.0; xmax = h;   # patch dimensions
              ymin =  0.0; ymax = h;
              height = h;

              xperiod = 8.0 * h;                      #  periodicity in -x
              yperiod = 4.0 * h;                      #  periodicity in -y
      }
      patch_2 {
              type = "box";                   # patch type: "box" || "hill"

              xmin = 4.0 * h ; xmax = 5.0 * h;   # patch dimensions
              ymin =  2.0 * h; ymax = 3.0 * h;
              height = h;

              xperiod = 8.0 * h;                  # periodicity in -x
              yperiod = 4.0 * h;                  # periodicity in -y
      }
  • EXP2 
    h = 16 # [m], building height
      patch_1 {
              type = "box";

              xmin = 0.0; xmax = h;   # patch dimensions
              ymin =  0.0; ymax = h;
              height = h;

              xperiod = 8.0 * h;                      #  periodicity in -x
              yperiod = 4.0 * h;                      #  periodicity in -y
      }
      patch_2 {
              type = "box";                   # patch type: "box" || "hill"

              xmin = 4.0 * h ; xmax = 5.0 * h;   # patch dimensions
              ymin =  2.0 * h; ymax = 3.0 * h;
              height = h/2.0;

              xperiod = 8.0 * h;                  # periodicity in -x
              yperiod = 4.0 * h;                  # periodicity in -y
      }
  • EXP3
h = 16 # [m], building height
        patch_1 {
                type = "box";

                xmin = 0.0; xmax = h/2.0;   # patch dimensions
                ymin =  0.0; ymax = h/2.0;
                height = h;

                xperiod = 8.0 * h;                      #  periodicity in -x
                yperiod = 4.0 * h;                      #  periodicity in -y
        }
        patch_2 {
                type = "box";                   # patch type: "box" || "hill"

                xmin = 4.0 * h ; xmax = 4.5 * h;   # patch dimensions
                ymin =  2.0 * h; ymax = 2.5 * h;
                height = h;

                xperiod = 8.0 * h;                  # periodicity in -x
                yperiod = 4.0 * h;                  # periodicity in -y
        }

Stratified idealized urban cases

coming soon

Realistic urban developments

cooming soon

Outputs

for output see notebooks here

Example of output for EXP1 image

Comparison