HYBRID LARGE EDDY SIMULATION/REYNOLDS AVERAGED NAVIER-STOKES FORMULATION FOR NUMERICAL WEATHER PREDICITON H. A. Hassan North Carolina State University,

Slides:



Advertisements
Similar presentations
Parametrization of surface fluxes: Outline
Advertisements

Parametrization of PBL outer layer Martin Köhler Overview of models Bulk models local K-closure K-profile closure TKE closure.
LARGE EDDY SIMULATION Chin-Hoh Moeng NCAR.
Introduction to Computational Fluid Dynamics
School of Mechanical, Aerospace & Civil Engineering Postgraduate Research Conference, PGR-MACE10 Performance of two k T -k L -ω models in a separation-induced.
F.Nimmo EART164 Spring 11 EART164: PLANETARY ATMOSPHERES Francis Nimmo.
TURBULENCE MODELING A Discussion on Different Techniques used in Turbulence Modeling -Reni Raju.
Turbulent Models.  DNS – Direct Numerical Simulation ◦ Solve the equations exactly ◦ Possible with today’s supercomputers ◦ Upside – very accurate if.
UNCLASSIFIED 3.5: Eddy Seeding for Improved WRF- LES Simulations Using Realistic Lateral Boundary Conditions Brian Gaudet, Aijun Deng, David Stauffer,
THE PARAMETERIZATION OF STABLE BOUNDARY LAYERS BASED ON CASES-99 Zbigniew Sorbjan Marquette University, Milwaukee Zbigniew Sorbjan Marquette University,
Direct numerical simulation study of a turbulent stably stratified air flow above the wavy water surface. O. A. Druzhinin, Y. I. Troitskaya Institute of.
0.1m 10 m 1 km Roughness Layer Surface Layer Planetary Boundary Layer Troposphere Stratosphere height The Atmospheric (or Planetary) Boundary Layer is.
Introduction to surface ocean modelling SOPRAN GOTM School Warnemünde: Hans Burchard Baltic Sea Research Institute Warnemünde, Germany.
D A C B z = 20m z=4m Homework Problem A cylindrical vessel of height H = 20 m is filled with water of density to a height of 4m. What is the pressure at:
Eddy Viscosity Model Jordanian-German Winter Academy February 5 th -11 th 2006 Participant Name : Eng. Tareq Salameh Mechanical Engineering Department.
Implementation of the quasi-normal scale elimination (QNSE) theory of turbulence in a weather prediction model HIRLAM Veniamin Perov¹, Boris Galperin².
Boundary Layer Meteorology Lecture 4 Turbulent Fluxes Energy Cascades Turbulence closures TKE Budgets.
Training course: boundary layer II Similarity theory: Outline Goals, Buckingham Pi Theorem and examples Surface layer (Monin Obukhov) similarity Asymptotic.
Estimation of Prandtls Mixing Length
LES of Turbulent Flows: Lecture 3 (ME EN )
Transport Equations for Turbulent Quantities
FUNDAMENTAL EQUATIONS, CONCEPTS AND IMPLEMENTATION
Basic Concepts of Numerical Weather Prediction 6 September 2012.
Introduction to Numerical Weather Prediction 4 September 2012.
The Air-Sea Momentum Exchange R.W. Stewart; 1973 Dahai Jeong - AMP.
CFD Modeling of Turbulent Flows
July 2001Zanjan, Iran1 Atmospheric Profilers Marc Sarazin (European Southern Observatory)
How to use CFD (RANS or LES) models for urban parameterizations – and the problem of averages Alberto Martilli CIEMAT Madrid, Spain Martilli, Exeter, 3-4.
AMBIENT AIR CONCENTRATION MODELING Types of Pollutant Sources Point Sources e.g., stacks or vents Area Sources e.g., landfills, ponds, storage piles Volume.
Page 1© Crown copyright Distribution of water vapour in the turbulent atmosphere Atmospheric phase correction for ALMA Alison Stirling John Richer & Richard.
1 LES of Turbulent Flows: Lecture 11 (ME EN ) Prof. Rob Stoll Department of Mechanical Engineering University of Utah Fall 2014.
Xin Xi. 1946: Obukhov Length, as a universal length scale for exchange processes in surface layer. 1954: Monin-Obukhov Similarity Theory, as a starting.
Comparison of Different Approaches NCAR Earth System Laboratory National Center for Atmospheric Research NCAR is Sponsored by NSF and this work is partially.
Comparison of convective boundary layer velocity spectra calculated from large eddy simulation and WRF model data Jeremy A. Gibbs and Evgeni Fedorovich.
Governing equations: Navier-Stokes equations, Two-dimensional shallow-water equations, Saint-Venant equations, compressible water hammer flow equations.
Budgets of second order moments for cloudy boundary layers 1 Systematische Untersuchung höherer statistischer Momente und ihrer Bilanzen 1 LES der atmosphärischen.
Mass Transfer Coefficient
Richard Rotunno NCAR *Based on:
Xin Xi Feb. 28. Basics  Convective entrainment : The buoyant thermals from the surface layer rise through the mixed layer, and penetrate (with enough.
10th AIAA/CEAS Aeroacoustics Conference, Manchester, UK, May, Refraction Corrections for Surface Integral Methods in Jet Aeroacoustics FongLoon.
A canopy model of mean winds through urban areas O. COCEAL and S. E. BELCHER University of Reading, UK.
LES of Turbulent Flows: Lecture 2 (ME EN )
Some Aspects of Parameterisation in Small Scaled Models Thomas Roschke German Weather Service Department of Aviation Meteorology.
Estimating the Optimal Location of a New Wind Farm based on Geospatial Information System Data Dec Chungwook Sim.
An ATD Model that Incorporates Uncertainty R. Ian Sykes Titan Research & Technology Div., Titan Corp. 50 Washington Road Princeton NJ OFCM Panel Session.
CITES 2005, Novosibirsk Modeling and Simulation of Global Structure of Urban Boundary Layer Kurbatskiy A. F. Institute of Theoretical and Applied Mechanics.
Further steps towards a scale separated turbulence scheme: Matthias Raschendorfer DWD Aim: General valid (consistent) description of sub grid scale (SGS)
Adaptive Optics in the VLT and ELT era Atmospheric Turbulence
NC STATE UNIVERSITY Prediction of Optical Scale Turbulence with a Typical NWP Code: Lessons Learned (?) Projected to Petascale Computing D. Scott McRae.
Numerical study of flow instability between two cylinders in 2D case V. V. Denisenko Institute for Aided Design RAS.
Convection in Flat Plate Boundary Layers P M V Subbarao Associate Professor Mechanical Engineering Department IIT Delhi A Universal Similarity Law ……
Aerospace Engineering N. C. State University Air Terminal Wake Vortex Simulation D. Scott McRae, Hassan A. Hassan N.C. State University 4 September 2003.
Page 1© Crown copyright Modelling the stable boundary layer and the role of land surface heterogeneity Anne McCabe, Bob Beare, Andy Brown EMS 2005.
Applied NWP [1.2] “…up until the 1960s, Richardson’s model initialization problem was circumvented by using a modified set of the primitive equations…”
Turbulence Models Validation in a Ventilated Room by a Wall Jet Guangyu Cao Laboratory of Heating, Ventilating and Air-Conditioning,
BOUNDARY LAYERS Zone of flow immediately in vicinity of boundary Motion of fluid is retarded by frictional resistance Boundary layer extends away from.
Parameterization of the Planetary Boundary Layer -NWP guidance Thor Erik Nordeng and Morten Køltzow NOMEK 2010 Oslo 19. – 23. April 2010.
1 LES of Turbulent Flows: Lecture 7 (ME EN ) Prof. Rob Stoll Department of Mechanical Engineering University of Utah Spring 2011.
Direct numerical simulation has to solve all the turbulence scales from the large eddies down to the smallest Kolmogorov scales. They are based on a three-dimensional.
A revised formulation of the COSMO surface-to-atmosphere transfer scheme Matthias Raschendorfer COSMO Offenbach 2009 Matthias Raschendorfer.
The Standard, RNG, and Realizable k- Models. The major differences in the models are as follows: the method of calculating turbulent viscosity the turbulent.
Interfacing Model Components CRTI RD Project Review Meeting Canadian Meteorological Centre August 22-23, 2006.
Development of the two-equation second-order turbulence-convection model (dry version): analytical formulation, single-column numerical results, and.
Pier Siebesma Today: “Dry” Atmospheric Convection
Numerical Investigation of Turbulent Flows Using k-epsilon
Overview of Downscaling
Turbulence closure problem
APISAT 2010 Sep. 13~15, 2010, Xi’An, China
*supported by the Army Research Office
transport equations for the scalar variances
Presentation transcript:

HYBRID LARGE EDDY SIMULATION/REYNOLDS AVERAGED NAVIER-STOKES FORMULATION FOR NUMERICAL WEATHER PREDICITON H. A. Hassan North Carolina State University, Raleigh, NC Presented at Workshop on Petascale Computing NCAR Mesa Laboratory 5-7 May 2008

Acknowledgment This research was supported in part by AF C-0019, Dr. George Jumper is the Technical Monitor, and a subcontract to NWRA, Dr. Joe Werne is the Technical Monitor.

1.Motivation for the Proposed Approach 2.Review of the Statistical Approaches for Describing Optical Turbulence 3.Proposed treatment of Optical turbulence 4.Comparison with experiment 5.Suggestions for improving Numerical Weather Prediction Codes 6.Concluding Remarks Presentation Outline

1.About five years go, we started an effort at North Carolina State University to improve predictions of Atmospheric Optical Turbulence (AOT) 2.AOT refers to turbulence in the atmosphere that causes fluctuations in the refractive index of air. 3.Primary sources of high altitude AOT are gravity waves and jet streams 4.Gravity waves arise from a number of sources including topography, convection and wind shear 5.Turbulence layers resulting from these gravity waves have thicknesses of the order of tens of meters and stretch many kilometers in the horizontal direction Motivation

6.Weather prediction codes, such as MMS, are used to currently provide velocity, temperature, relative humidity, etc. for calculating the index of refraction structure function. Typical grids in such codes are in the range 4-30 Km in the horizontal direction and 0.3 – 1Km in the vertical direction with some further refinement in the planetary boundary layer 7.As is seen from the above, mesoscale codes present a number of problems: a.Resolution is not adequate to resolve sufficiently gravity waves and optical turbulence b.Mesoscale turbulence parameterization is suited for Large Eddy Simulation (LES). However, mesoscale model resolution are not resolving LES scales 8.Parameterization used to calculate Cn 2 is based on rather old turbulence models Motivation Cont’d

1.For wavelength near the visible range, C n is given by (1) where P is the pressure in Pascals, T and  are the temperature and potential temperature in Kelvin 2.(2) where C  is the dissipation rate of the variance of potential temperature, and  is the dissipation rate of the variance of velocity or turbulent kinetic energy (TKE) Review of Statistical Approaches

3.Assuming a Richardson number of zero, setting production equal to dissipation, assuming turbulent eddy viscosity to be equal to turbulent diffusivity (i.e., Prandtl number of one), and using mixing length theory, the following expression for is obtained (3) where L 0 is the outer length scale and z is the vertical direction. Current statistical models calculate L 0 in terms of one or more of the following: z,  T/  z,  u/  z using available experimental data. Statistical Approaches

1.In order to address resolution issues, a grid adaptation approach is employed. My colleague Dr. Scott McRae has already discussed this topic 2.Instead of using mixing length theory and other simplifying assumptions to derive an expression for, model equations were derived for   and  (variance of vorticity or enstrophy) =  /, with being the kinematic viscosity 3.Rather than employ ad hoc empirical equations for   and , equations were derived from the exact Navier-Stokes equations and then parameterized. Similar approach was used to derive a TKE and an equation for the variance of potential temperature. Proposed Approach

4.Advantage of above procedure is that one ends up with a set of equations that are tensorially consistent, Galilean invariant, coordinate system independent and contain all relevant physics 5.Because grid resolution is such that LES cannot resolve all scales, a Reynolds Averaged Navier-Stokes (RANS) formulation is seamlessly blended with the LES formulation. This approach ensures that contribution of all scales are taken into consideration Proposed Approach Con’d

6.As an illustration the TKE has the form Proposed Approach Con’d

Comparison with Experiment

Figure 1. Mesh spacing in the vertical direction

Figure 2. Comparison of u Profiles, Vandenberg Case, Balloon 1

Figure 3. Comparison of Profiles, Vandenberg Case, Balloon 1

Figure 4. Comparison of Profiles, Vandenberg Case, Balloon 1

Figure 5. Comparison of u Profiles, Holloman Case

Figure 6. Comparison of v Profiles, Holloman Case

Figure 7. Comparison of Profiles, Holloman Case

Suggestions for Improving NWP Codes 1.Every effort should be made to reduce damping in numerical schemes to the absolute minimum. Moreover, damping should be scaled such that it is reduced when the grid is refined 2.There is a need to improve surface boundary conditions to reflect more accurately the shape and roughness of the torrain.of the terrain 3.There is a need to improve or replace the sponge boundary condition 4.LES cannot resolve all scales. Therefore, there will always be a need to add a RANS component to account for unresolved scales. In this case additional blending functions should be explored.

Suggestions for Improving NWP Codes 5. Some consideration should be given to developing a version of WRF by utilizing preconditioning. The net effect of this is to provide various filtering of acoustic modes depending on local conditions.

Conclusions 1.In spite of the damping that exists in MM5, turbulence models based on the Navier-Stokes equations provide a viable alternative to ad hoc turbulence models and should be used in current and future NWP codes 2.Current model is capable of predicting optical turbulence if numerical damping can be reduced and finer grids are employed in NWP codes.