Land shallow cumulus convection (ARM)

Case description

The ARM Cumulus (ARMCU) case simulates a diurnal cycle of shallow cumulus convection over the ARM (Atmospheric Radiation Measurement) Southern Great Plains site. This is a classic case for testing cumulus parameterization and LES modeling of continental shallow convection. The case exists in two configurations: 1D (single column) and 3D (LES).

Warning

The 3D configuration requires MPI parallelization and the 1D configuration can only be run on a single core.

Configuration

Parameter

1D (ARMCU_1D_CONDSAMP)

3D (ARMCU_LES)

Category

Idealized cases

HPC cases

Dynamics

3D with 1D turbulence

3D LES

Horizontal grid spacing

40000 m (1x1)

500 m (64x64)

Integration length

54000 s (15 hours)

43200 s (12 hours)

Time step

100 s

2 s

Coriolis effect

enabled

enabled

Turbulence

TKEL (1D)

TKEL (3D)

Cloud scheme

ICE3

ICE3

Deep convection

NONE

NONE

Shallow convection

EDKF

NONE

Radiation

NONE

NONE

LES diagnostics

enabled

enabled

Declination

Configuration

Turbulence

Grid

ARMCU_1D_CONDSAMP

TKEL (BL89)

1x1

ARMCU_LES/HM21

TKEL (HM21)

64x64

ARMCU_LES/DEAR

TKEL (DEAR)

64x64

Steps

Step

Script

001_prep_ideal

run_prep_ideal_case_xyz / run_prep_ideal

002_mesonh

run_mesonh_xyz / run_mesonh

Specificities

Scientific specificities

  • Idealized forcing using time-varying surface fluxes (heat, moisture, momentum)

  • Surface latent heat flux switches from positive to negative during the transition from daytime heating to nighttime cooling

  • Deep large-scale forcing from ARM field campaign data

  • Conditional sampling (CONDSAMP) enabled with 3 sample levels for cloud statistics

  • LES diagnostics with time-averaging from 3600s to 43200s (12h daytime period)

Technical specificities

  • 1D: Cartesian flat domain with single point (NIMAX=1, NJMAX=1), 100 vertical levels

  • 3D: 64x64 horizontal grid, 100 vertical levels with 40m spacing

  • Model top at 1000m

  • Initialization from ARM SGP sounding (June 21, 1997)

  • Cyclic boundary conditions for 3D

Validation

  • Vertical profiles of mean quantities: MEAN_RC, MEAN_RR, MEAN_U, MEAN_V, MEAN_W, MEAN_THL

  • Subgrid flux profiles: SBG_WTHL, SBG_WRT

  • Total water and mass flux: MEAN_RT, MEAN_MF

  • Cloud fraction validation at different levels

Numerical ressources

1D: Single CPU (1 node, 1 core) 3D: 2 nodes, 128 cores (MPI parallel)

References

  • Brown, A.R., et al. (1999): Large-Eddy Simulation of Atmospheric Boundary Layer, BLM, 91, 271-289.

  • Xu, K.M., and D.K. Araker (1995): The behavior of macrophage-like cells in small cumulus clouds. J. Atmos. Sci., 52, 2923-2942.