next up previous
Next: TLM Validation Up: Automatic Differentiation as a Previous: Methodology - Sensitivity to

4. 3-D Storm Model and Control Run

Our experiments are performed using the sensitivity-enhanced code generated from Version 4.0 of the ARPS, which is three dimensional, fully compressible, and nonhydrostatic. The prognostic variables, solved on the Arakawa C grid [1], include Cartesian velocity components, perturbations of potential temperature and pressure, mixing ratios of water vapor, cloud water, and rain water, and turbulent kinetic energy. The advective modes are computed on large timesteps with a leap-frog time scheme and second-order centered space differencing, whereas the acoustic modes are integrated on small timesteps with an implicit scheme. Kessler-explicit warm-rain microphysics is employed [12]. An extensive description of the model can be found in the ARPS users guide [22].

The computational domain consists of tex2html_wrap_inline482 grids in the horizontal with a grid size of 1 km. In the vertical, a stretched grid system is employed for 35 grids with a resolution of 150 m near the ground and 850 m at the top of the model domain. The model is run for 140 min, with a large timestep of 6 sec and a small timestep of 1 sec. The detailed model configuration for our experiments is described in [15].

The simulation to verify the computation of derivatives by ADIFOR is made by using the HALF4 (supercell) hodograph and thermodynamic sounding from [7], the latter of which has a surface mixing ratio of 15 g/kg. This wind profile consists of a semi-circular arc of 10 m/s radius that turns clockwise over the lowest 4 km starting with the surface easterly winds. The (westerly) wind is constant, with height above 4 km at a speed of 10 m/s. The convection is initiated by a 4 K thermal perturbation placed in the boundary layer. The simulated supercell develops rapidly during the first 30 minutes and becomes quasi-steady thereafter, with a sustained updraft of around 47 m/s. In Figure 1, the surface outflow boundary velocity and vertical velocity at 4 km are depicted for t = 50 and 120 min. The storm moves to the west initially and then turns northeastward as it grows in vertical extent, forming a strong surface cold pool.

  figure1
Figure 1: Control simulation: vertical velocity at 4 km (positive in solid and negative in dashed lines at an interval of 2.5 m/s) and perturbation potential temperature at surface (dotted lines with contours larger than -2.0 K at an interval of 0.5 K) at (a) 50 min and (b) 120 min


Another storm is triggered by convergence along the northern gust front. Also, as the northeast part of the gust front intensifies, a new cell develops along it (t = 50 min; Figure 1a), constituting three distinct cells. As the northern part of the gust front moves northward out of the model domain by 60 min, the two northern storms decay, and other weak secondary storms mill around the northern lateral boundary. The dominant storm thereafter is the isolated supercell, which travels southeastward along the leading edge of the expanding cold pool. A secondary storm develops northeast of the main storm after 100 min (Figure 1b), merging into the main storm by 140 min.


next up previous
Next: TLM Validation Up: Automatic Differentiation as a Previous: Methodology - Sensitivity to

Seon Ki Park
Sun Nov 17 12:46:57 CST 1996