Joe Klemp - PowerPoint PPT Presentation


PPT – Joe Klemp PowerPoint presentation | free to download - id: 64e12d-YzU0M


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation

Joe Klemp


Title: PowerPoint Presentation Author: Joe Klemp Last modified by: ccsm Created Date: 6/18/2007 8:41:39 PM Document presentation format: On-screen Show (4:3) – PowerPoint PPT presentation

Number of Views:25
Avg rating:3.0/5.0
Slides: 29
Provided by: JoeK72
Learn more at:


Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: Joe Klemp

WRF-CAM Collaborations for Future Weather and
Climate Applications
3 km WRF/ARW 25 h reflectivity forecast
CCSM2 20 year annually averaged precipitable water
Joe Klemp National Center for Atmospheric
Research Boulder, Colorado, USA
WRF User Participation
4/4/08 Registered Users WRF
Principal Partners 250 U.S.
Universities (135) 1257 U.S.
Government Labs 388 Private Sector
629 Foreign
------- Total
6439 New V3.0
registrations 1158

7597 Foreign countries represented 103
V3.0 release
3200 active subscribers to Curre
ntly averaging 400 email inquiries per month to
WRF/ARW Model Overview
  • Characteristics, Features, Capabilities
  • Nonhydrostatic dynamical solver using higher
    order numerics and conservative prognostic
  • Data-assimilation options through WRF-Var
  • Flexible, extensible to range of WRF applications
  • Parallel, efficient on range of computers in WRF
  • Movable, feature following nested grids
  • Coupling to other models

WRF/HYCOM Coupling through ESMF
WRF Data Assimilation WRF-Var
WRF-Var is a unified variational data
assimilation system built within the WRF software
framework, used for application in both research
and operational environments.
AFWA Worldwide Theatres
Unifying elements
  • Domains Regional/global
  • Techniques 3D-Var, 4D-Var, Hybrid Var/Ensemble
  • Code Single code for research, development and
    release. Supported by NCAR/MMM.
  • Software Engineering WRF framework
  • Model Runs with WRF, and also KMA global model
  • Radiative transfer Common model (CTRM) with JCSDA

KMA T213/426 Global
Used operationally at AFWA, Korea/KMA, China/BMB,
Taiwan/CWB Antarctica/AMPS, UAE
WRF-ARW Nonhydrostatic Dynamic Core
  • Terrain-following hydrostatic pressure vertical
  • Arakawa C-grid
  • 3rd order Runge-Kutta split-explicit time
    differencing, 5th or 6th order
    differencing for advection
  • Conserves mass, momentum, dry entropy, and
    scalars using flux form prognostic equations
  • Minimal additional computational damping

Observed Kinetic Energy Spectra
WRF-ARW Kinetic Energy Spectra
Physics Options Implemented in WRF
  • Microphysics Kessler-type (no-ice), Lin, Goddard,
  • WSM3/5/6, Ferrier, Thompson, Morrison
  • Cumulus Convection New/Old Kain-Fritsch, Grell
  • Betts-Miller-Janjic, Grell-3
  • Shortwave Radiation Dudhia (MM5), Goddard,
  • Longwave Radiation RRTM, GFDL, CAM
  • Turbulence Prognostic TKE,
  • Smagorinsky, constant diffusion
  • Surface Layer Similarity theory, MYJ
  • Land-Surface 5-layer soil model, RUC LSM
  • Noah unified LSM, CLM

In progress
Model Physics in High Resolution NWP
Physics No Mans Land
Resolved Convection
Cumulus Parameterization
3-D Radiation
Two Stream Radiation
PBL Parameterization
WRF-ARW Real-Time Convective Forecasts
Year Horizontal Grid Domain PBL Microphysics Land-Surface
2003 4 km 2000 x 2000 km YSU Lin (5 cat) OSU
2004 4 km 2800 x 2600 km YSU Lin (5 cat) OSU
2005 4 km 3900 x 3000 km YSU WSM6 (6 cat) Noah
2006 4 km 3900 x 3000 km MYJ WSM6 (6 cat) Noah
2007 3 km 3330 x 2760 km MYJ Thompson (6 cat) Noah
3 km WRF-ARW Forecast 2007 NOAA HWT Spring
Forecast and composite radar reflectivity for
tornadic squall line at 01 UTC 4/14/07
25 h WRF/ARW 3 km forecast
2 km NOWRAD Mosaic
2005 Real-time 4 km ARW Moving-Grid Hurricane
Katrina Wind Forecast, Initialized 00Z 27 Aug 2005
Number of cases
Atmospheric Chemistry Applications WRF-Chem
Current features
  • Online coupling of dynamics
  • and chemistry
  • Automatic generation of chemical reactions and
    constituents through Kinetic Pre-Processor
  • Aqueous phase chemistry coupled to microphysics
    and aerosol schemes
  • Photolysis coupled with hydrometeors, aerosols
    and convective parameterizations
  • Biogenic and anthropogenic emissions
  • Feedback direct effect of aerosols on short-wave

Two week comparison of 36 h WRF-Chem 27 km
forecasts with AIRNOW daily 8-h max O3
Real-time AQ forecasts with WRF-Chem since 2002
Improvement in 36 h WRF-Chem 27 km forecasts of
daily 8-h max O3
The Nested Regional Climate Model Phase 1
1996-2000 2000-2005 Tropical Simulations
Tropical Channel, 36 km, N/S boundaries 1-way
nested into NCEP Reanalysis with specified SST,
Kain-Fristch Cu Parameterization, CAM radiation
and YSU boundary layer.
45o N
30o S
4 km nested domain inside 12 km and 36 km
domains, fully 2-way interactive, Dudhia cloud
physics, no cumulus parameterization.
Wind Speed (m/s)
Monsoon Activity
Monsoon intraseasonal activity is evident even in
the unfiltered OLR data (Compare the outlined
Northward propagation of the convective anomalies
Low-Pressure system moves from the Bay of Bengal
into the Continent
NRCM Hurricane Simulations
North Atlantic 12 and 4 km 2-way nested
domains. Run may-October 2005
4 km res. domain
12 km res. domain
No cumulus parameterization in 4 km domain
North Atlantic and North American Regional
Climate Changes
  • The goal is to simulate the effects of climate
    change on precipitation across the intermountain
    West States and tropical cyclones, with a focus
    on the Gulf of Mexico.
  • Outer domain nested into CCSM, which also
    provides surface conditions.
  • 1996-2000, then three 5-y time slices out to 2050
  • 5 ensemble members for each period
  • Allocated half of NCAR IBM Blue Fire for 2 months

CAM physics - WRF status
  • CAM 3 radiation - In WRF Version 3.0 release
  • CLM 3 land-surface - Mostly added to WRF
    non-repository code
  • - requires linking to WRF initial state and
    land properties
  • - implemented directly into WRF executable
    code rather than by coupling
  • PBL - potentially Holtslag Boville or Bretherton
    scheme could be added to WRF with little effort
  • Convection - several CAM options, but considering
    Neale-Richter, probably little effort
  • Cloud microphysics - 2-moment Morrison scheme in
    CAM is being considered (complications include
    whether to also add sub-column or cloud fraction
    parameterizations to WRF)
  • Subgrid-scale dynamics - potentially gravity wave
    drag, but currently porting Hong orographic drag
    scheme instead. This is already in progress.

Porting CAM physics to WRF is a high priority !
WRF Global Model
Global WRF on a lat-long grid
  • Adapted from community development at Cal Tech
    for planetary atmospheres
  • Functional system for nested nonhydrostatic
    global simulations
  • Baseline for future nonhydrostatic global model

10 day precipitable water forecast, initialized
7-11-2007 12Z
810 x 405 x 41 (x,y,z), 50 km grid at the
equator, 200 second timestep
Towards a Next Generation Climate-Weather-Earth
System Model
  • The Atmospheric Component

(Courtesy of Morris Weisman)
  • Existing and future applications require
    meso-scale and cloud-scale resolution in a global
  • Current climate models are poor weather models,
    and current weather models are poor climate
  • Opportunity to leverage the diverse interests and
    experience of the climate and weather communities
    to create and share a next-generation atmospheric
    simulation system.

What is Wrong With Our Existing Global Models?
  • They do not scale to 104 - 105 processors. (e.g.
    lat-long grid models)
  • They were not constructed for mesoscale/cloudscale
    applications (e.g. physics, numerics, tuning).
  • Why use higher resolution?
  • Explicitly simulate convective systems
  • Capture system evolution (growth, decay,
  • Resolve moisture redistribution, cloud systems.
  • Remove need for deep cumulus parameterization
    (with sufficient resolution - ?x lt a few km).
  • Explicitly simulate gravity waves, wave breaking
  • Remove the need for gravity-wave drag
  • Better resolution of external forcing
  • topography, land-use, etc.

Towards a Next Generation Climate-Weather-Earth
System Model
The Atmospheric Component
  • Objectives
  • For weather, climate, and ESM requirements
  • Construct an atmospheric dynamical core that runs
    efficiently on existing and future MPP computers,
    and has sufficient flexibility for diverse
  • Unify and share physics, and leverage development
    efforts, where possible.

ESM Atmospheric Component WG
  • Phil Rasch, Dan Marsh, Bill Skamarock (co-chairs)
  • Atmospheric Dynamical Core
  • Subgroup
  • Bill Skamarock (facilitator)
  • 25 members (14 outside NCAR)
  • Fall workshop being planned
  • Atmospheric model physics
  • Subgroup
  • Phil Rasch and Dan Marsh (facilitators)
  • Still being formed
  • CCSM and WRF scientists

The working (sub)groups serve as forums for
discussions and information dissemination, and
provide recommendations, critiques, and reviews.
ESM Atmospheric Core WG Members
  • George Bryan
  • Joe Klemp
  • Peter Lauritzen
  • Ram Nair
  • Phil Rasch
  • Amik St-Cyr
  • Bill Skamarock
  • Piotr Smolarkiewicz
  • Joe Tribbia
  • Henry Tufo
  • Dave Williamson

Joern Behrens (Alfred Wegener
Institute) Jean-Michel Campin (MIT) Phil Colella
(Lawrence Berkeley Labs (LBL)) Bill Collins
(LBL) John Drake (ORNL) Dale Durran (Univ.
Washington) Christiane Jablonowski (Univ.
Michigan) Jin Lee (NOAA/GSD) Bill Putnam
(NASA) Todd Ringler (LANL) Richard Rood (U.
Mich.) Mark Taylor (Sandia) John Thuburn
(Exeter Univ.) Robert Walko (Duke Univ.)
Features of a Weather and Climate ESM Dynamical
  • Strong Consensus
  • Fully compressible nonhydrostatic equations
  • Mass conserving
  • Scalar mass conserving, consistent.
  • Positive-definite (PD) transport for PD scalars
  • Local refinement capability
  • Regional modeling capability

Features of a Weather and Climate ESM Dynamical
  • Desirable Features
  • Monotonic transport options
  • Horizontal grid uniformity
  • (little variation in cell area)
  • Horizontal grid isotropy (dx dy)
  • Energy conservation?

Features of a Weather and Climate ESM Dynamical
  • Evaluation Metrics and Issues
  • Good energetics
  • KE spectra (resolving both synoptic- and
  • Dissipational and frictional heating
  • Efficient (cost for a given accuracy level)
  • On MPP architectures
  • Global to cloud scales, weather and climate
  • Low-order schemes or higher-order schemes?
  • Implicit or explicit formulations?
  • The grid should be invisible

Some Possible Discretizations for the Sphere
lat-long grid
triangular grid
hexagonal grid
cubed sphere
yin-yang grid
  • High-order solvers
  • Grid-cell isotropy
  • Uniformity of grid-cell sizes
  • Grid continuity, handling of special points
  • Local refinement

Nonhydrostatic Modeling on Hexagonal C-Grid
  • New ESM core must function effectively at both
    global and cloud-resolving scales
  • Icosahedral hexagonal grid has good potential to
    satisfy these requirements.
  • Hexagonal grids have not been evaluated for
    cloud-scale applications.
  • C-grid staggering on hexagonal grids not widely

12 pentagons
Global Hexagonal Grid
Limited Area Hexagonal C-Grid
3-D Cloud Model on Hexagonal Grid
Splitting Supercell at 2 hours
  • Research Progress
  • Constructed a 3-D limited-area hexagonal-grid
    cloud model (based on WRF/ARW numerics) to
    evaluate performance.
  • Documented that hexagonal-grid cloud simulations
    are at least as accurate and computationally more
    efficient than those on a conventional
    rectangular grid.

1 km Hexagonal Grid Simulation
Atmospheric Component The Path Forward
  • 3 years from now
  • Capable of global cloud-resolving simulations -
    days to months.
  • This implies
  • Developing and testing of promising dynamical
    core prototypes, and a decision on a first ESM
    dynamical core within 2 years.
  • Developing and porting of select physics.
  • Developing a computational framework suitable for
    the dynamical solver, the physics, and
    model-coupling needs.
  • Developing archival capabilities, and analysis
    and post-processing tools.
  • Resources for development, testing, and near-term