Numerical Modeling at NCEP/EMC:

Progress and Plans

Geoff DiMegoMesoscale Modeling BranchGeoff.DiMego@noaa.gov

301-763-8000 ext722130 November 2010

NCEP

T O P I C S• Operational Model Suite Overview• Central Computer System (CCS) Acquisition Status• Gridpoint Statistical Interpolation (GSI)• Global Forecast System (GFS)• HiResWindow• NOAA Environmental Modeling System (NEMS)• North American Mesoscale (NAM)• NARRE, HRRRE• SREF, VSREF• RTMA, Analysis of Record

MostlySatellite+Radar

North AmericanMesoscale

NMM

North American Ensemble Forecast System

Hurricane GFDL

HWRF (NMM)

GlobalForecastSystem

Rapid RefreshGSD’s RUC moving toWRF-ARW for RR

ClimateCFS

~3B+ Obs / Day

Short-RangeEnsemble Forecast

The NCEP Operational Model SuiteItems in RED are Meso Branch Concerns

MOM3Coupled

Global DataAssimilation

Dispersion, Ash,Smoke & DustARL’s HYSPLIT

OceansHYCOM

WaveWatch III

Air QualityNAM +EPA’s CMAQ

NMM + ARW[+SREF = HREF]

Severe Weather

Regional DataAssimilation

NMM + ARW +ETA + RSM

GFS, Canadian Global Model, FNMOC +

VSREFRUC + NAM

DGEXNMM

R T M ANOAH Land Surface Model

Fire Weather : NMM

CCS Acquisition Status• Draft Request for Proposals (RFP) was posted to FedBizOpps 12 Nov 2011

• Highlights (& LOWLIGHTS!)– 10 year contract: 5-yr Base, 3-yr option & 2-yr transition option

– Vendors must provide two facilities• One must be at least 120 miles from DC• Both facilities must be at least 120 miles apart

– Initial delivery will be the same size as the current CCS• No increase in compute capacity of system or disk storage

• Additional funding is being pursued, but has not yet been identified

– Any system architecture will be considered• Unix (vendor flavor) or Linux based

• Homogeneous or heterogeneous

• Current contract has been extended for 2 years (FY12+FY13) for transition– One year to bring the two facilities on-line

– One year for transition

– Provides for NO increase in compute capacity or storage

CCS Acquisition Status• Remaining Schedule

– Vendor responses to draft RFP – 30 November 2010– Industry Day – 6 December 2010

– Revise & Post final RFP – ASAP– Vendors submit responses– Proposals evaluated & Set Competitive Range– Provide remaining vendors comments & questions– Receive Best and Final Offers (BAFO)– Review BAFO & make recommendation

– Award Contract – no later than 30 September 2011– New contract starts – No later than 1 October 2011– System Operational – No later than 1 October 2013

Gridpoint Statistical Interpolation

• Multi-Agency development effort led by NCEP/EMC– ESRL / GSD + ESRL / PSD

– Joint Center for Satellite Data Assimilation• NOAA, NESDIS, NASA, DOD

– Code management (SubVersion) with regression testing

• Community supported via DTC in Boulder– GSI Community Tutorial June 28-30, 2010

• Includes hybrid approach: Ensemble Kalman-Filter + 3D- or 4D-VAR– GSI already had 4D-VAR capability added by Y.Tremolet/R.Todling of NASA/GMAO

with many solution options for both 3D- and 4D-VARiational

– First hybrid version now installed and tested in GSI

– Works for global & NEMS/NMMB on full or subset of model domain / resolution

– Reads external ensemble(s) of different resolution

– Testing dual resolution to make the hybrid affordable, connecting NMMB/NDAS to operational GEFS & experimental EnKF

GSI UpgradesGlobal upgrade ~March 2011 Faster code (~9%), improved optimization

and additional optionsRecomputed background errorsLimit moisture to be >= 1.e-10 in each outer

iteration and at the end of analysisLocate buoys at 10 m (from 20 m)Ambiguous vector qc for ASCAT dataSatellite radiance related changes

Update to radiative transfer model - CRTM 2.0.2Inclusion of Field of View Size/Shape/Power for

radiative transferRelax AMSU-A Channel 5 QCRemove down weighting of collocated radiancesInclusion of uniform (higher resolution) thinning

for satellite radiances

Stratospheric satelliteImproved OMI QC Removal of redundant SBUV/2 total ozoneRetune SBUV/2 ozone ob errorsInclusion of SBUV from NOAA-19

New ob sources for NAM ~May 2011

New conventional obs– MESONET ps, T, q

– ACARS moisture (WVSS-II)

– MAP Profiler winds– RASS Profiler Tv

– WINDSAT & ASCAT ocean winds (from scatterometer)

New unconventional obs– Satellite Radiances

• AMSUA from aqua & NOAA19

• HIRS4 & MHS from NOAA19

• IASI from METOP-A

– Refractivity• GPS radio occultation

7

8

GFS Resolution Increase (July 2010)

• Horizontal Resolution Increase from T382 to T574 spectral or from ~35 km to ~27 km

• Physics Upgrades & Enhancements– Radiation and Cloud– Boundary Layer– Mass flux Shallow Convection– Deep Convection– Removal of Negative Water Vapor

• Gravity-Wave Drag Parameterization• Better Hurricane Relocation & Snow analysis

9

• Modified GFS shallow/deep convection and PBL– Detrainment from all levels (deep convection)– PBL diffusion in inversion layers reduced (decrease erosion of marine stratus)

• Eliminates grid-point precipitation “bombs”

Impact of GFS Physics Upgrade Implemented 28 July 2010

Observed Operational GFS Upgraded Physics GFS

24 h accumulated precip ending 12 UTC 14 July 2009

T574

T382

ECMWF

10

Low Wind Speed Bias in Stratosphere Will Be Corrected with Reduced Diffusion in ~March 2011 GSI+GFS-Bugfix Bundle

T574 with reduced diffusion

11

GFS FY2012 Upgrade Plans Model not yet been finalized - current candidates are:

T-574 / L64 Eulerian Sub-grid scale cloud scheme using Monte-Carlo Independent Column Approximation Ferrier Microphysics (from NAM) Surface Roughness over water NSST Model – produces realistic diurnal cycle for SST

T-878 / L64 Semi-Lagrangian Longer time step permitted for dynamics & advection Uses identical grid as T-574 to perform physics Recently solved the noise issue at the top of the model Additional efficiency has been gained - greater resolution may be possible

Include some/all of the Eulerian changes

Aerosols using GOCART (with NASA) WAFS products harmonized with UK Met (WAFC-London)

Offline GFS-GOCART 7-day Simulation of Iceland’s Eyjafjallajökull Volcano

• Daily 48 Hours Forecasts made from April 14 2010 to April 20 2010, CPU~ 3 hours clock each day

• Driven by operational GFS (T-382 remapped to 1 deg x 1 deg)• 5 size bins (radius) DU1: 0.1 – 1.0 µm, DU2: 1.0 - 1.8 µm, DU3: 1.8 – 3.0 µm, DU4: 3.0 – 6.0 µm, DU5: 6.0 – 10.0 µm

• Continuous release of 1x106 kg/hr in 1x1 grid box at ~5 km level for each bin, total=5x106 kg/hr

• Results shown:– 0-24 hour forecast from April 14 to April 20, 24-48 hour forecast for last day (April 21 01Z- April 22 -00Z)

– Hourly average total column concentration, sum of all 5 bin (du1 + du2 + du3 + du4 + du5)

Guam

18Z

06Z

00Z12Z

00Z,12Z 06Z,

18Z00Z,12Z

4.0 km WRF-NMM

5.15 km WRF-ARW

48 hr fcsts from both

Unless there are hurricanes

Expanded PR domain

~Feb 2011 Upgrade of HiResWindow• Upgrade NMM & ARW to WRF v3.2

• Use improved passive advection in both cores

• Add High Resolution Ensemble Forecast (HREF) & BUFR product generation

• Daily displays of these runs can be seen at:http://www.nco.ncep.noaa.gov/pmb/nwprod/analysis/ andhttp://www.emc.ncep.noaa.gov/mmb/mmbpll/nestpage/

SREF mean of 21 members 24 h total qpf

HREF mean of 44 members 24 h total qpf

ARW 24 h total qpf

NMM 24 h total qpf

20101020/06f48 20101020/06f48

20101020/03f51 20101020/06f48

NMM ARW

Sample BUFR output from HiresW planned 2011 Upgrade

NEMS Component StructureMAIN

EARTH(1:NM)

Ocean Atm IceBelow the dashed line, the source codes are organized by the model developers.

FIM

Dyn Phy Wrt

NEMS

Ensemble Coupler

GFS

Dyn Phy Wrt

NMM

Dyn Phy

Domains(1:ND)

Wrt

All boxes represent ESMF components (now supported by NOAA/ESRL). NEMS

LAYER

ARW

Dyn Phy Wrt

WRF-ChemGOCART

CMAQ

NEMS NMM-B Nests

- Static, 1-way interaction.

For Spring 2011 implementation: ●

- Unique sets of processors assigned to each domain to optimize the balance of work and minimize the clocktime required.

- Boundaries fed by the parent every parent timestep during the integration.

Future●

- Moving, 1-way interaction; now under development.

- Any parent-nest ratio (integer) can be used.

- Nest is ‘grid-associated’ with its parent.

- Static/Moving, 2-way interaction. Flexibility may be restricted in domain processor assignment and parent-nest ratios.

- Free-standing nests that are not ‘grid-associated’ with parents.

19Z 11 August

Tropical Depression 5

1000 m REFD, 10 m winds

Gulf Spill 1.33 km NEMS/NMMB nest

19

~May 2011 NAM Upgrade

Current NAM• WRF-NMM (E-grid)• GSI analysis• 4/Day = 6 hr update• Forecasts to 84 hours• 12 km horizontal• 12 hr pre-forecast

assimilation period with 3hr updates (catch-up)

New NAM• NEMS based NMMB• B-grid replaces E-grid• Parent remains 12 km to 84 hr• Multiple Nests Run to 60 hr

– 4 km CONUS nest– 6 km Alaska nest– 3 km HI & PR nests

• Single locatable ~1.33-1.5 km Single locatable ~1.33-1.5 km FireWeather/IMET/DHS run to 36hrFireWeather/IMET/DHS run to 36hr

Microphysics Development• Added WRF’s WSM6 microphysics (W. Wang)

– Tests over 4-km CONUS (conus4) and 4-km CONUS in 12-km AQ domain runs (nest)

– Similar skill in both runs, WSM6 ~30% slower

• Advection of condensate (W. Wang)

– Added “spec_adv” flag to NMMB• Advect total condensate only (false)

• Advect individual hydrometeors (true)

– Works for wsm6 & fer microphysics

– Similar skill seen for either advection option

• Concluded no benefit to compensate for greater expense

Physics Testing: Reduced Convective Triggering in 4 km CONUS nest vs 12 km Parent

0-60 h Cu QPF

Parent (BMJ)

Parent (BMJ)

Nest (BMJ_DEV)

Nest (BMJ_DEV)

Deep Cu Cloud TopPressure

(hPa)

22

Turned on gravity wave drag/mountain blocking in 6 km Alaska nest to reduce

10-m wind bias

Last run w/o GWD First run w/GWD

Microphysics Changes Impacts the Simulated Radar Reflectivities

New ferrier

• Higher composite dBZ in new vs. ops fer • Output is from 4-km CONUS nest runs

but interpolated to 12-km grid 218.

Ops ferrier

Parent & Nest Reflectivity Loop

Left: 12-km “Air Quality” Domain (~1/2 size of current NAM)- Control BMJ convection (same as in NAM)- Modified fer microphysics

Right: 4-km CONUS domain- BMJ_DEV convection (reduced triggering)- Modified fer microphysics

New Cloud Fraction(reduced for cold, high clouds)

GOES W Vis GOES W IR

NAM Total Cloud Fraction (%) New Total Cloud Fraction (%)

Hurricane Earl near Puerto Rico

12 km NMMB parent 3 km Puerto Rico nest

May 2010 27/16Z to 28/12Z

MD Backdoor Coldfront in 1.33km Nest

May 2010 27/16Z to 28/12Z

MD Backdoor Coldfront in 1.33km Nest

Convergence of NAM & RUC into hourly NARRE & HRRRE• There is a signed agreement on NARRE between NCEP/EMC and ESRL/GSD • Based on NEMS common modeling infrastructure• Ensembles:

• Sample uncertainty within membership• Initial & Lateral Boundary conditions• Dynamics & Physics

•Provide full description of uncertainty

•Can adapt to rapidly evolving science of underlying data assimilation and modeling

2011NAM

• NEMS based NMMB• Bgrid replaces Egrid• Parent remains at 12 km to 84 hr• Multiple Nests Run to 60hr

– 4 km CONUS nest– 6 km Alaska nest– 3 km HI & PR nests

• Reinstate Fire Weather/IMET Support/DHS run to 36hr– Locate a single 1.33-1.5 km1.33-1.5 km run– In either CONUS or Alaska

Rapid Refresh• WRF-based ARW• NCEP’s GSI analysis• Expanded 13 km Domain

to include Alaska• Experimental 3 km HRRR

RUC-13 CONUS domain

WRF-Rapid Refresh domain – 2010

Original CONUS domain

Experimental 3 km HRRR

2014-2015North American Rapid Refresh

ENSEMBLE (NARRE)• NMMB (from NCEP) & ARW (from ESRL) dynamic cores• Common use of NEMS infrastructure and GSI analysis• Common NAM parent domain at 10-12 km• Initially ~6 member ensemble made up of equal numbers of

NMMB- & ARW-based configurations• Hourly updated with forecasts to 24 hours• NMMB & ARW control data assimilation cycles with 3

hour pre-forecast period (catch-up) with hourly updating• NAM & SREF 84 hr forecasts are extensions of the 00z,

06z, 12z, & 18z runs – for continuity sake.– SREF will be at same 10-12 km resolution as NARRE by then– SREF will have 21 members plus 6 from NARRE for total of 27

• NARRE requires an increase in current HPCC funding

34

2015-2016High Resolution Rapid Refresh

ENSEMBLE (HRRRE)• Each member of NARRE contains 3 km nests

– CONUS, Alaska, Hawaii & Puerto Rico/Hispaniola nests– The two control runs initialized with radar data & other hi res obs

• This capability puts NWS/NCEP[+OAR/ESRL] in a position to – Provide NextGen Enroute AND Terminal guidance (FWIS-like)– Provide PROBABILITY guidance with full Probability Density

Function specified, hence uncertainty information too– Provide a vehicle to improve assimilation capabilities using hybrid

(EnsKF+4DVar) technique with current & future radar & satellite– Address Warn-on-Forecast as resolutions evolve towards ~1 km

• NAM nests are extensions of the 00z, 06z, 12z & 18Z runs.

• HRRRE requires an increase in current HPCC funding over and above that required for the NARRE

35

2015-2017A Catch-Up Cycle for NARRE & HRRRE

could constitute the Analysis of Record• Catch-up = reach back in time to include late arriving obs• Assimilate ALL in situ and remote data sources • Use state-of-the-art 4-dimensional data assimilation

technique– Likely a hybrid of Ensemble Kalman-Filter and 3D-/4D-Variational– Able to take quick advantage of its evolution

• Use state-of-the-art nonhydrostatic numerical models– Advanced Research WRF (ARW) core from NCAR & ESRL/GSD– Non-hydrostatic Multiscale Model on B-Grid (NMMB) from NCEP– Interoperable physics from WRF community & NCEP operations– Able to take quick advantage of their evolution

• Extend to include NextGen required parameters• This AoR requires an increase in current HPCC funding

Review SREF Implementation (Oct. 27th 2009)- Jun Du et al.

• Upgrade model versions – WRF-NMM from v2.0+ to v2.2+– WRF-ARW from v2.0+ to v2.2+– RSM from v2007 to v2009

• Increase horizontal resolution– WRF-NMM from 40km to 32km– WRF-ARW from 45km to 35km– RSM from 45km to 32km

• Adjust membership– Replace 2 Eta (BMJ-sat) members with 2 WRF-NMM members– Replace 2 Eta (KF-det) members with 2 WRF-ARW members

• Enhancement physics diversity of RSM: replace Zhao cloud scheme with Ferrier cloud scheme for 3 SAS members

• Enhance initial perturbation diversity: Replace regional bred perturbations with global ET perturbations for 10 WRF members

Next SREF Implementation Plan (Q4FY2011) - Geoff DiMego and Jun Du

Models and configurations• Drop Eta (6 members) and RSM (5 members)• Add 7 NEMS-NMMB members and 2 each to WRF members,

yielding• Membership and resolution of future SREF will be

– 7 NEMS-NMMB at 22 km (from 32 km)– 7 WRF-NMM at 22 km (from 32 km)– 7 WRF-ARW at 25 km (from 35 km)

• Initialization diversity of future SREF will be– 7 NEMS-NMMB from NAM / NDAS + ETR– 7 WRF-NMM from GFS / GDAS + ETR– 7 WRF-ARW from Rapid Refresh + ETR

• Ensemble job scripts have been generalized to run both WRF & NEMS members significantly simplifying future implementations

• Output same fields as current operational SREF

Post process

• Precipitation calibration

SREF Implementation Plan for FY11-14 - Geoff DiMego and Jun Du

North America Ensemble Forecast System’s extension to regional ensemble (NAEFS_LAM)

• Based on 5th NAEFS conference: May 17-19th, 2010• Need - Both U.S. and Canada need to run high-resolution regional ensembles for high-impact weather. • Benefit - More resource can be spent on increasing model resolution but not on increasing ensemble membership as

well as increasing forecast diversity.• Canadian REPS: 20 members with GEM, downscaled GEFS IC perturbations, 30km, 48hr, NA domain, later 2010

implementation (B08RDP, VC2010, Haiti earth quake relief effort) (21+20=41 member SREF)• Problem - Both countries are big in domain and don’t have enough computing resources to run such a high-

resolution ensemble with large enough ensemble size.• Evolving Plans for NAEFS_LAM

– 2010: verification and implementation of REPS at Canadian side– 2011/12: research to see benefit of combining NCEP SREF and CMC REPS and resolve technical details about the data/variable

exchange following NAEFS– 2013: operational implementation at both centers

Others

• Stochastic physics– Convective parameterization of Teixeira et al - NRL

• Ensemble transform with rescaling (ETR) initial perturbations– Consistent with boundary perturbations from GEFS

• Resolution– Looking for higher resolution

06 09 1200 0318 211206

NAM RUC

• NAM cycles always older than RUC VSREF gives more weight to RUC

• Example: Ensemble member combination for 06Z cycle run

4 NAM cycles, weighted 0.7, 0.5, 0.3, 0.1, respectively

6 RUC cycles, weighted 1.0, 0.9, 0.8, 0.7, 0.6, 0.5, respectively

Forecast hour extended to 12 hr (with extension of RUC forecasts to 18hr)

15 18

06Z cycle: VSREF’s ensemble member configuration

21 00

Very Short Range Ensemble Forecast (VSREF) System [courtesy of Binbin Zhou]

VSREF Webpage

Download grids from here

Google Map of 4 RTMA DomainsFirst Phase of Analysis of Record

Real Time Mesoscale Analysis

Analyzed every hour on the NWS’ NDFD grids10 m wind + est. anal. uncertainty

2m Temperature + est.anal.unc.

2m dew point + est.anal.unc.

Sfc pressure + est.anal.uncertainty1 hr precip (Stage 2)GOES Eff. Cloud Amount

Courtesy of Yan Zheng University of Utah

42

2.5 km (top) vs 5 km (bottom)

NAM 12 hr Forecast Ri-Based PBL Height with Verifying RAOBs

Initial Test RTMA Analysis of PBL Height