Abstract:

The dataset is a collection of idealized atmospheric simulations; the fluxes and vertical gradients of tracers are used to estimate the eddy-diffusivity for a variety of stability regimes (stable, neutral, convective). A Large Eddy Simulation (LES) model generates vertical profiles of fluxes which allow the exploration of a new approach to specifying diffusive and non-diffusive contributions to the vertical eddy diffusivity without specifying shape or scale. Diffusivities, fluxes, and shapes are provided for surface and entrainment dominated fluxes; those results and the bulk ratios of diffusive to total fluxes and velocity scales are compared to previously-explored velocity scales. This dataset supports the publication: Chor, T., McWilliams, J. C., & Chamecki, M. (2020). Diffusive-nondiffusive flux decompositions in atmospheric boundary layers. Journal of the Atmospheric Sciences, 1–55. doi:10.1175/jas-d-20-0093.1

Suggested Citation:

Chor, Tomas. 2020. Dataset for: Diffusive-nondiffusive flux decompositions in atmospheric boundary layers. Distributed by: GRIIDC, Harte Research Institute, Texas A&M University–Corpus Christi. doi:10.7266/SD6S0Y9M

Publications:

Chor, T., McWilliams, J. C., & Chamecki, M. (2020). Diffusive-nondiffusive flux decompositions in atmospheric boundary layers. Journal of the Atmospheric Sciences, 1–55. doi:10.1175/jas-d-20-0093.1

Purpose:

To investigate a new approach to specify eddy diffusivity profiles in models of the atmospheric boundary layer by estimating diffusive and nondiffusive fluxes without a prior assumption of shape or scaling.

Data Parameters and Units:

The dataset consists of a README file; “overall_atm.nc”, which details the ratio of nondiffusive to total fluxes and the ratio of bulk velocity scale to bulk velocity scale in Troen and Mahrt, 1983; and four files presenting profiles of fluxes, diffusivities, and shape functions for the four different values of stability (000 = neutral, 015 = rolls, -020 = stable, and 376 = plumes.) The profile data files have the parameters wc (vertical turbulent flux of tracers, [m/s]); Fd_opt (turbulent diffusive flux of tracers, [m/s]); Fnd_op (turbulent nondiffusive flux of tracers, [m/s]); K_opt (eddy diffusivity from optimized method, [m^2/s]); K_SFT (surface-forced tracer eddy diffusivity, [m^2/s]); K_EFT (entrainment-forced tracer eddy diffusivity, [m^2/s]); K_grad (eddy diffusivity from TM86, [m^2/s]); Gs_opt (surface-driven nondiffusive flux shape function from optimization, [dimensionless]); Ge_opt (entrainment-driven nondiffusive flux shape function from optimization, [dimensionless]); Gs_d66 (surface-driven nondiffusive flux shape function when Ge=0, [dimensionless]) ; Gs_EDMF (surface-driven nondiffusive flux shape function from EDMF, [dimensionless]); tracer [“SFT”, “EFT”]; z/z_T (depth normalized by the depth of the top of the entrainment layer, [dimensionless]); Λ (stability parameter, w*^3/u*^3 = -κ z_e / L_o, [dimensionless]); wcs (tracer flux at surface, [m/s]); wce (tracer flux at entrainment, [m/s]); u_star_meas (measured friction velocity, [m/s]);w_star_meas (measured convective velocity, [m/s]); ze_meas (measured height of maximum entrainment, [m]); zt_meas (measured top of entrainment layer", [m]); L_o (Monin-Obukhov length, [m]). The bulk ratios file has the parameters tracer [“SFT”, “EFT”]; opt [“unconstr”, “constr”]; Λ (stability parameter, w_^3/u_^3, [nondimensional]), R_F (ratio of nondiffusive to total fluxes, [nondimensional]), R_U (ratio of optimized velocity scales to predicted by Troen and Mahrt, 1986).

Methods:

Large-eddy simulation (LES) model-produced vertical fluxes of passive tracers and temperature were used to explore the diffusive/nondiffusive decomposition of vertical eddy diffusivity profiles. Four simulations are presented with varying stability regimes (neutral, stable, plumes (convective with predominantly plume structures), and rolls (convective with predominantly rolls structures). The LES model has 256 grid points in the horizontal direction, 400 gridpoints in the vertical direction, a sponge layer in the top one-quarter of the domain to provide an open boundary, and periodic horizontal boundary conditions. Stability differences were primarily produced by the surface buoyancy forcing, and the resolution was varied in order to resolve the eddies responsible for vertical mixing. A constant 5 m/s geostrophic velocity was prescribed in the x-direction, and inertial oscillations were minimized. The two tracers modeled were a surface-forced passive tracer and an entrainment-forced passive tracer. The diffusive flux was split into surface-forced tracer (SFT) and entrainment-forced tracer (EFT) components acting on different concentrations and different initial and boundary conditions. The SciPY package optimization library was used to determine the optimal forms of the three flux components. Further details may be found in the publication.

Provenance and Historical References:

Troen, I. B., & Mahrt, L. (1986). A simple model of the atmospheric boundary layer; sensitivity to surface evaporation. Boundary-Layer Meteorology, 37(1-2), 129–148. doi:10.1007/bf00122760

Individual Files: