Case 4: Deep convection, supercell
Part I: Idealized 3-D supercell event
This case is intended to simulate a tornadic supercell observed on 29 June 2000 northwest of the town of Goodland, Kansas. The event was documented in Kuhlman et al. (2006) and the purpose of this workshop case is investigation of sensitivities of microphysics to observed storm features and evolution. Kuhlman et al. (2006) produced a reasonable simulation of the event using an idealized sounding and warm bubble perturbation, and we will replicate many of their steps to model the observed storms. Comparisons among model simulations and to detailed observations will be the primary focus of the workshop.
We request that modelers adhere to as many specifications for setting up the model as is practical but understand that some model differences are likely. Please create a model domain that is minimum 80 x 80 km2 in the horizontal and 20 km in the vertical with maximum grid spacing of 1000 m in the horizontal. If possible, we recommend using a stretched vertical coordinate that produces a spacing of approximately 100 m or less near the ground and approximately 500 m above 6.5 km. The model should be integrated for 3 hours with a timestep no longer than 5 seconds or less if needed for numerical stability. If possible, we advise using a Rayleigh dampening layer at model top with damping coefficient (inverse damping timescale) of 0.025 s−1 within the model's top 5 km to damp spurious waves in the stratosphere. The upper and lower boundaries are free-slip with zero vertical velocity and lateral boundary conditions are open. Surface fluxes should be zero, and radiative transfer and Coriolis effect should be neglected for this experiment. Horizontal and vertical turbulent diffusion may be decided by each researcher, and case leaders will utilize a 1.5 order TKE scheme (Skamarock et al. 2007). Case leaders employed the "Box method" for scalar advection and "Crowley scheme" for momentum within the Straka Atmospheric Model.
The model is initialized with the environmental temperature, moisture, and wind profiles used by Kuhlman et al. (2006), which was based on a research rawinsonde launched at 2022 UTC from Goodland, KS. The initial convective available potential energy (CAPE) is 2875 J kg−1. Convection is initially triggered by adding a thermal with maximum perturbation in potential temperature of 3 K centered at a height of 1.5 km and varying as the cosine squared to the 0 K perturbation at its edge. The thermal has a horizontal radius of 9 km and a vertical radius of 1.5 km and should be placed at horizontal grid point (x=35, y=30) if the suggested grid spacing and points are used. Please adjust to match if you use different grid spacing. If possible, translate the grid with the mean storm motion using the vector (ugrid=11.0, vgrid=−1.75) m s−1. If a moving grid is not possible, expand number of grid point to capture the storm for at least 2.5 hours.
We request specific graphics be prepared for the workshop to facilitate comparisons among all participants. Please bring these graphics at minimum, but you are welcome to bring additional plots of model output that you find illustrative. Please return to this section later for the requested graphics list because we are still developing the material (last updated 18 Mar 2008).
Additional data/files discussed in the description document
The NCAR S-Pol radar collected volume scans on this storm and the files below can be used
to compare model results. More details on the contents of these files is found within the link above.
Questions should be directed to Matthew Gilmore.
*NOTE: The particle type within the netcdf files below were improperly interpolated
giving non integer values. Corrected netCDF files containing only the particle identification (using
nearest neighbor) are now available separately from
this UIUC web page.
- Goodland NEXRAD animation (1.6 MB Quicktime movie)
- SPol data v150.nc.gz (13.0 MB gzipped netCDF file)
- SPol data v154.nc.gz (95.6 MB gzipped netCDF file)
- SPol data v158.nc.gz (98.1 MB gzipped netCDF file)
- SPol data v160.nc.gz (106 MB gzipped netCDF file)
- SPol data v163.nc.gz (103 MB gzipped netCDF file)
- SPol data v165.nc.gz (103 MB gzipped netCDF file)
- SPol data v168.nc.gz (93.1 MB gzipped netCDF file)
- SPol data v170.nc.gz (102 MB gzipped netCDF file)
- SPol data v173.nc.gz (93.0 MB gzipped netCDF file)
- SPol data v175.nc.gz (88.0 MB gzipped netCDF file)
- SPol data v154.v5d.gz (73.6 MB gzipped Vis5d file)
- SPol data v158.v5d.gz (75.0 MB gzipped Vis5d file)
- SPol data v160.v5d.gz (76.1 MB gzipped Vis5d file)
- SPol data v163.v5d.gz (76.9 MB gzipped Vis5d file)
- SPol data v165.v5d.gz (77.3 MB gzipped Vis5d file)
- SPol data v168.v5d.gz (68.0 MB gzipped Vis5d file)
- SPol data v170.v5d.gz (74.9 MB gzipped Vis5d file)
- SPol data v173.v5d.gz (68.1 MB gzipped Vis5d file)
- SPol data v175.v5d.gz (66.3 MB gzipped Vis5d file)
- SPol data steps.v5d.gz (656 MB gzipped Vis5d file)
Part II: Case study: 29 Jun 2000 from STEP experiment
For those modelers wishing to obtain gridded analyses and observed surface and rawinsonde data for this case in order to attempt a retrospective model forecast, please contact Greg Thompson. If there is enough interest, links will be added within this page to these data.
References
Kuhlman, K. M., C. L. Ziegler, E. R. Mansell, D. R. MacGorman, and J. Straka, 2006: Numerically simulated electrification and lightning of the 29 June 2000 STEPS supercell storm. Mon. Wea. Rev., 134, 2734–2757.