Presentation is loading. Please wait.

Presentation is loading. Please wait.

19 December 2012 1 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Examining the Sensitivity of.

Similar presentations


Presentation on theme: "19 December 2012 1 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Examining the Sensitivity of."— Presentation transcript:

1 19 December 2012 1 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Examining the Sensitivity of an Idealized Model to Changes in the Initial Potential Temperature Perturbation Chip Helms

2 19 December 2012 2 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius The Idealized Model: CM1 Created by George Bryan (NCAR) 3D, non-hydrostatic, non-linear, cloud-resolving, idealized model –No data assimilation –Uses a horizontally constant field for the base state –Adds perturbations to base state e.g. warm bubble, cold blob, forced convergence Benefits of using CM1 –Conserves mass and energy better than other modern cloud models –Faster and uses less memory than other models for idealized studies –Very flexible, can be used for a large variety of studies

3 19 December 2012 3 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Goal Qualitatively examine CM1 sensitivity to: –Initial maximum magnitude of θ' –Initial horizontal warm bubble radius Potential implications of study: –Sensitivity of cloud-resolving models to temperature anomalies e.g. magnitude and extent of an urban heat island

4 19 December 2012 4 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Methodology Two sets of runs to test sensitivities

5 19 December 2012 5 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Model Settings

6 19 December 2012 6 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Sensitivity to magnitude of θ' Composite Reflectivity Cell split occurs earlier as θ' increases

7 19 December 2012 7 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Sensitivity to Warm Bubble Radius Cell splitting is related to the interaction between vorticity and updrafts Cell split occurs earlier as radius increases Look at updraft strength evolution

8 19 December 2012 8 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius 0.5 K Delay in reaching peak updraft strength is non- linear function of θ' 1.0 K1.5 K2.0 K

9 19 December 2012 9 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Relatively low sensitivity at 7 km radius and above Relatively high sensitivity at 6 km radius and below Model Dispersion 2-6Δx  2-6 km radii

10 19 December 2012 10 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Conclusions As θ' or warm bubble radius increases, storm cell splits earlier and reaches peak updraft strength earlier –More sensitive to θ' than radius Small warm bubble radius runs have resolution and dispersion issues –Impacts 5 km radius most –Very little sensitivity at or above 7 km –Suggests sensitivity is due to model dispersion

11 19 December 2012 11 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Questions

12 19 December 2012 12 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Peak Updraft Limit Function of CAPE CAPE = 1946  w max = 62 m/s Actual < 55 m/s Difference due to assumptions of CAPE

13 19 December 2012 13 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Initial Background Conditions Hodograph corresponds to these levels

14 19 December 2012 14 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Hodograph Refresher Trace of wind vector direction (azimuth) and magnitude (radius) Straight hodographs suggest cells will split Veering (backing) hodographs suggest right (left) cell will be dominant

15 19 December 2012 15 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Model Settings - Continued Boundary conditions: –Open-radiative lateral boundary conditions –Zero-flux top/bottom boundary conditions Not included in these runs: –Atmospheric radiation –Surface drag –Surface fluxes

16 19 December 2012 16 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Sensitivity to magnitude of θ' bfds

17 19 December 2012 17 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Outlying Model Runs θ' = 0.5K lags behind other runs Radius = 5 km lags significantly behind other runs –Possibly due to turbulent mixing having a greater impact on tight gradients smaller size of anomaly would be diffused faster –Could also be due to dampening near the 2Δx scale (recall Δx = 2 km)

18 19 December 2012 18 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Detailed model settings 5 th order horizontal/vertical advection schemes Negative moisture is corrected by taking moisture from adjacent grid cells No additional artificial diffusion beyond subgrid turbulence scheme Sixth order diffusion scheme (coefficient = 0.040) TKE subgrid turbulence scheme Zero flux boundary condition for vertical diffusion of winds/scalars at top/bottom of domain Uses Rayleigh damping at upper levels (e-fold time = 1/300, applied above 14km), but not near horizontal boundaries Uses Klemp-Wilhelmson time-splitting, vertically implicit pressure solver (as in MM5, ARPS, WRF), coeff for divergence damper = 0.10, slight foreward-in-time bias used for vertically implicit acoustic solver (alpha = 0.60) Moisture scheme: Morrison double-moment scheme –Hail is used for large ice category –cloud droplet concentration: 250 cm^-3 (marine=100,continental=300) No Coriolis force Includes dissipative heating No energy fallout term Open-radiative lateral boundary scheme: Durran-Klemp (1983) formulation Initial base-state sounding: Weisman-Klemp analytic sounding Initial base-state wind profile: RKW-type profile Initial pressure perturbation is zero everywhere

19 19 December 2012 19 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Sensitivity to magnitude of θ' Composite Reflectivity Cell split occurs earlier as θ' increases Cell splitting is related to the interaction between vorticity and updrafts


Download ppt "19 December 2012 1 ConclusionsResultsMethodologyBackground Chip HelmsSensitivity of CM1 to Initial θ' Magnitude and Radius Examining the Sensitivity of."

Similar presentations


Ads by Google