1

 

William. M. LapentaDirector, National Centers for Environmental Prediction

NOAA/NWSGeoff DiMego , John Derber , Yuejian Zhu , Hendrik Tolman , Vijay Tallapragada,

Shrinivas Moorthi , Mike Ek , Mark Iredell, Stan Benjamin

NCEP

The NOAA Operational Numerical Guidance Suite:

Major Upgrade Plans for FY14 Prepared by the NCEP Environmental Modeling Center

 

1. Global Observing System

2. Computers (WCOSS, AWIPS2)

3. Data Assimilation, Modeling, Ensembles and Post Processing

Three Major Components of the Numerical

Prediction Enterprise….

Everything you read, see or hear about weather, climate and ocean forecasts is

based on numerical prediction

2

NOAA Operational Data Assimilation and Modeling

NOAA Science Serving Society….

NOAA Operational Numerical Guidance Supports the Agency Mission

– Numerical Weather Prediction at NOAARequired for agency to meet service-based metrics

– National Weather Service GPRA* Metrics

(* Government Performance & Results Act)Hurricane Track and Intensity Winter Storm WarningPrecipitation Threat Flood WarningMarine Wind Speed and Wave Height

– Operational numerical guidance:Foundational tools used by government, public and private

industry to improve public safety, quality of life and make business decisions that drive U.S. economic growth

3

Lead Time and

Accuracy!

Regional Hurricane

GFDLWRF-NMM

WRF(ARW, NMM)NMMB

Climate ForecastSystem (CFS)

Short-RangeEnsemble Forecast

NOAA’s Operational Numerical Guidance Suite (Jan 2014)

GFS, MOM4,NOAH, Sea Ice

North American Ensemble Forecast System

GEFS, Canadian Global Model

Dispersion HYSPLIT

Air QualityCMAQ

Regional NAMNMMBNOAH

3D-V

ARD

A

Regional Bays•Great Lakes (POM)

•N Gulf of Mexico (FVCOM)•Columbia R. (SELFE)•Chesapeake (ROMS)

•Tampa (ROMS)•Delaware (ROMS)

SpaceWeather

ENLIL4

North American Land Surface Data Assimilation

SystemNOAH Land Surface Model

Global SpectralNOAH3D

-En-

Var

DA

Global Forecast System (GFS)

3D-V

ARD

A

3D-V

ARD

A

WRF ARW

Rapid Refresh

3D-V

ARD

A

WavesWaveWatch III

Ocean HYCOM

Ecosystem EwE

Global Ensemble Forecast System (GEFS)21 GFS Members

ESTOFSADCIRC

SURGESLOSH

P-SURGESLOSH

WRF ARW3D-V

AR

DA

High Resolution RRNEMS Aerosol Global Component (NGAC)

GFS & GOCART

WRF(ARW, NMM) & NMMB

High Res Windows

5

24-h Cycle 01 February 2014

Numerical Guidance Suite Execution WCOSS NOAA Supercomputer

Nu

mb

er

of

No

de

s

Time of Day (UTC)

00 06 12 18 00CFS NAM GFSGEFS SREF

HyCOM

SREF Para

RAPv2 Para

RAP

NRC Recommendation I.b:NRC Recommendation I.b:

Numerical Weather Prediction:“The National Weather Service (NWS) global and regional numerical weather prediction systems should be of the highest quality and accuracy, with improvements driven by user needs and scientific advances.

To achieve this goal, the NWS should give priority to upgrading its data assimilation system and increasing the resolution of its deterministic and ensemble modeling systems.”

Can’t forget physics and Post Processing

7

The NOAA Modeling Strategy…High Level Perspective

Moving away from the “model of the day” Continue to pursue multi-model approach to ensembles

• Don’t forget: ensemble systems only as good as the modeling system it is built from

Priorities for deterministic development are clear:1. Data assimilation (methodology and observations)2. Model physics

Why do we continue to underplay this important part of the enterprise?

Clouds, microphysics, radiation, land, ocean, ice, aerosols….includes coupling

3. Resolution—horizontal and vertical4. Dynamic core

Must consider advanced HPC technologies but don’t forget about the science

Regional systems shift to convection permitting applications NOAA encouraged to consider unified modeling approach

(UCACN)

 

Infrastructure/DA Upgrades

Physics upgradesCombined (best of all upgrades)

H14A H14BNest motion

(H140)

NOAH LSM

(H141)

Upgraded Ferrier (H142)

RRTMG (H143)

Ocean(H144)

 H214 (proposed

2014 HWRF)

Description

1. Sat Da with 61 levels, 2mb top2. Extended d02/d033. Upgraded vortex init.4. GSI upgrades

1. No Sat DA2. Include Invests in cycling

1. New nest motion & diagnostics

NOAH LSM

Separate species and F_rime advection with other upgrades

RadiationMPI-POM with new coupler

Baseline + physics + python based scripts*need to do test runs with new GFS in WCOSS

Cases

Whole 2011-2013 & selected 2008-2010 storms

Whole 2011-2013 & selected 2008-2010 storms

Priority casesPriority cases

Priority cases

Priority cases

Priority cases

Whole 2011-2013 & selected 2008-2010 storms

Due date Feb. 15 Feb. 15 Feb. 15 Feb. 15 Feb. 15 Feb. 15 Feb. 15 March 31

Platform Jet/WCOSS Jet/WCOSS Jet Jet Jet Jet Jet Jet/WCOSS

2014 HWRF pre-implementation test plan2014 HWRF pre-implementation test plan

Continuing on the same path of extensive testing of individual upgrades and their combination for multiple seasons that resulted in higher confidence in the model. Joint NHC/EMC collaboration during the pre-implementation testing. 8

9

Integrate Research Community into the Model development and Transition Proces

EMC Change Control Board•Scientific Integrity•Product Quality

•EMC Mgmt Approval•ACCOUNTABILITY

Implementation Phase• SPA’s build NCO parallel

from RFC’s•30-day NCO parallel

Test code stabilityTest dataflow

Products to NCEP Centers and EMC code developers•NCEP Centers

Evaluate impactAssessments to NCEP OD

R&D and Pre-Implementation Phase

•30-day NCO parallel stable•NCEP centers approve

•ACCOUNTABILITY•Briefing to NCEP Director

for final approval•ACCOUNTABILITY

Implementation•Generate RFC’s

•Submit RFC’s to NCO

Systematic

Systematic

Testing

Testing

10

Hurricane Forecast Improvement Project

Scope:•Improve hurricane forecast system/global forecast system to reduce error in intensity and track•Make better use of existing observing systems; define requirements for future systems to enhance research and operations capabilities and impacts•Expand and improve forecaster tools and applications to add value to model guidance

Vision•Organize the hurricane community to dramatically improve numerical forecast guidance to NHC in 5-10 years

Goals•Reduce numerical forecast errors in track and intensity by 20% in 5 years, 50% in 10 years•Extend forecast guidance to 7 days with today’s skill at 5 days•Increase probability of detecting rapid intensification at day 1 to 90% and 60% at day 5

11

Operational: Systematic assessment and testing with well defined annual development cyclesHolistic end-to-end system testedEstablished metrics and baselinesCustomers included in process—final approvalLeads to consistent incremental improvement

Research:Development based on process oriented approach driven by funding program prioritiesOften event driven (case studies)Baseline: a control run that is not a baselineStudies emphasize limited changes in the systemNo independent customer feedback--metrics

Model Development:Incremental is not a “Bad” Word…

• Operational systematic end-to-end testing

• Increasing the gap between HWRF and AHW for skill and computational efficiency

• Annual incremental improvements produce revolutionary results

2010-2012 Atlantic Basin Intensity Error (Kts)

Forecast Hour

12

Forcing Factors Shaping NOAA Operational Numerical Guidance

Emerging Requirements Weather Ready Nation High impact events Weather to climate—seamless suite of guidance and products

Science and Technology Advances Observing systems High performance computing Data dissemination Numerical Guidance Systems

Data assimilation (methodology) Modeling (physics, coupling & dynamics) Ensembles (construction—initialization, membership, etc.) Intelligent post processing

Predictability convective systems Seasonal to interannual

13

Major System Upgrades (Planned)Q1-FY14 to Q1-FY15

• List does not include NOS or MDL systems

FY15

Q1 Q2 Q3 Q4 Q1

North American Ensemble Forecast System NAEFS X

Global Data Assimilation/Forecast System GDAS/GFS X

Global Forecast Ensemble System GEFS XNCEP Global Aerosol NGACNorth American Model NAM X X

High-Resolution Window HiRESW X

Short Range Ensemble System SREF X XRapid Refresh RAP X XHigh-Resolution Rapid Refresh HRRR X

Real Time Mesoscale Analysis RTMA X X

HYSPLIT HYSPLIT X

Hurricane Weather Research and Forecast HWRF X

GFDL Hurricane GFDL X

North American Land Data Assimilation NLDAS X

System

FY14

Real Time Ocean Forecast System X X(2)Wave Watch III X

Acronym

RTOFSWaves

XX X

SREF Interim Upgrade PackageSystem Configuration Modifications

To Be Implemented March 2014

Initial conditions: a)replace GFS land states (too moist) with NDAS land states in NMM & ARW members; b)correct inadvertent use of global initial conditions (too moist) with use of RAP for ARW members to reduce the surface wet and cold biases

NOAH LSM: •eliminate negative soil moisture fractions for NMM and ARW members;•eliminate “urban swamp” (causing too cold surface temperature over urban regions during heat wave periods) for NMMB members

GFS physics: 2 NMMB members to produce compatible cloud & ceiling guidance with the rest of SREF members

Post-Processor: remove use of snow in diagnosing cloud base height

Mapping: (eastward shift) in NMM member’s pressure-grib output files

14

Observed 36hr (00z, Sept. 12 – 12z, Sept. 13, 2013) accumulated precipitation.

SREF Prob > 100mm (~4”)/36hr

52hr

81hr 75hr 69hr

63hr 58hr

81hr 75hr 69hr

63hr 58hr 52hr

SREF Prob > 150mm (~6”)/36hr

Application to high-impact weather:

2013 Boulder Flood

SREF consistently indicating a historical heavy rain event over Boulder, CO region days ahead (Hamill 2014; plots from Rich

Grumm)

System Current Q4FY14 FY18

GDAS 80 member @ T574 Eulerian (55 km)

80 member @ T574 SL (35 km)

4DHybrid 80 member @ T1148 SL (17 km)

Analysis @ T574 Eulerian (27 km) using T254 Eulerian ensembles

Analysis Increment @ T574 SL (35 km) using T574 SL ensembles

Analysis Increment @T1148 SL (17 km) using T1148 SL ensembles

64 Vertical Levels 64 Vertical Levels 128 Vertical Levels

Uses GFS model below Additional Obs., Improved radiative transfer, many smaller changes, uses GFS model below

Additional Obs., Cloudy Radiances, Improved QC and ob. Errors, Ensemble Hurricane relocation, uses GFS model below

GFS T574 Eulerian (27 km) to 7.5 days

T1534 SemiLagrangian (13 km) to 10 days

T2000 SemiLagrangian (10 km) to 10 days

T254 Eulerian (55 km) days 7.5 to 16

T574 SL (35 km) days 10 to 16

T1148 SL (17 km) days 10 to 16

64 Vertical Levels 64 Vertical Levels 128 Vertical Levels

Enhanced Physics Higher top, non-hydrostatic, NEMS, Coupled Ocean, enhanced physics

Global Forecast System Evolution:2014 to 2018

Global Ensemble Forecast System (GEFS)

• Where will we get the biggest bang for the buck?• Membership• Horizontal resolution• Vertical resolution • Perturbations

• Are reforecasts a new emerging requirement? • Will require extensive testing

18

ED-4DVAR: T254L64 Experiments

• 3DVAR: 2x100 iterations

• Hybrid 3D EnVar: 80 member T126L46 Ens., fully coupled EnKF update, TLNMC on total increment, 75% ensemble & 25% static

• Hybrid 4D EnVar: 1x150 iteration, TLNMC on all time levels, Hourly TC relocation

19

NAM• Implemented 18 October 2011

• NEMS based NMM• Parent remains at 12 km• Multiple Nests Run to ~48hr

– ~4 km CONUS nest– ~6 km Alaska nest– ~3 km HI & PR nests– ~~1.5-2km DHS/FireWeather/IMET

possible

Rapid Refresh• Scheduled implementation 20 March 2012

• WRF-based ARW• Use of GSI analysisUse of 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

NCEP Mesoscale Modelingfor CONUS:

Verification from 1 October 2013 – 15 January 201424-h QPF scores(24-60 h Forecasts)

Pre

ssu

re (

mb

)

Vector Wind RMS (m/s)12km CONUS

Day 1 = BlackDay 2 = RedDay 3 = Blue

• Physics Modifications:• GWD/mountain blocking; more responsive to

sub-grid scale terrain variability

• BMJ convection : moister convective profiles, convection triggers less, increase 12 km bias

• RRTM radiation, latest version

• Ferrier-Aligo microphysics, tuned to improve severe storm depiction

• Improved snow depth algorithm in LSM

Test Results for NAM FY14 Upgrade Package

• Data Assimilation Modifications:• Hybrid variational-ensemble analysis with global

EnKF

• New satellite bias correction algorithm (same as in FY14 global upgrade)

• Cloud/radar assimilation in NDAS (planned)20

Ops

Parallel

6 8 10 12

.01 .10 .50.25 1.0.75 2.01.5 3.0 4.0

Precipitation Threshold (Inches)

BIAS

Equitable Threat

.01 .10 .50.25 1.0.75 2.01.5 3.0 4.0

Precipitation Threshold (Inches)

12-km Ops12-km Para4-km Ops4-km Para

12-km Ops12-km Para4-km Ops4-km Para

4

Rapid Refresh and HRRRNOAA hourly updated models

13km Rapid Refresh (RAP)

(mesoscale)

3km HRRR (storm-scale)

High-Resolution Rapid Refresh Scheduled NCEPImplementation 15 July 2014

Version 2 – scheduled NCEP implementationQ2 (25 Feb 14)

RAP

HRRR

Courtesy of Stan Benjamin 21

DataAssimilation

AssimilationType

PBL Pseudo Innovations

Soil Adjustment

Low-Cloud Building

Snow Cover Building

Lightning Obs(radar refl proxy)

Tower/Nacelle/Sodar Obs

RAPv1 GSI 3D-VAR No No No No No No

RAPv2 GSI EnKF-3DVARHybrid Yes Yes Yes Yes + rev

trimming Yes Yes

Model Version Horiz/VertAdvection

ScalarAdvection

Upper-LevelDamping

SW RadiationUpdate Land Use LSM PBL Microphysics Radiation

LW/SW

RAPv1 WRF-ARW V3.2+ 5th/3rd Monotonic w-Rayleigh

0.02 30 min USGS RUC 6-lev MYJ Thompson

V3.1RRTM/

Goddard

RAPv2 WRF-ARWv3.4.1+ 5th/5th Positive-

Definitew-Rayleigh

0.2 10 min MODISFractional

RUC 9-lev MYNN Thompson

v3.4.1RRTM/

Goddard

Major changes for RAPv2 Assimilation •hybrid ensemble assimilation•soil/PBL/surface/cloud assimilation

Model – - land-surface, PBL, cloud physics

RAPv2 enhancements from RAPv1

Courtesy of Stan Benjamin 22

Summary: RAPv2 vs. RAPv1

RMS vector wind errorVs. raobsSep-Nov 2013

Winds -- Consistent RAPv2 improvement, for both upper-air and surface

Moisture -- Improved RAPv2 RH forecasts aloft. Reduced RMS 2m dewpoint error, more so at night

Temperature -- Similar RAPv2 skill for upper-air. Improved skill for surface in warm seasonPrecipitation and clouds – Improvement for precipitation, slight improvement for ceiling

RAPv1 vs. RAPv2

RAPv2better

RAPv1better

Courtesy of Stan Benjamin 23

Model Version Assimilation Radar DA RadiationLW/SW Microphysics Cumulus

Param PBL LSM

RAP WRF-ARW v3.4.1+

GSI Hybrid 3D-VAR/Ensemble 13-km DFI RRTM/

GoddardThompson

v3.4.1 G3 + Shallow MYNN RUC 9-lev

HRRR WRF-ARWv3.4.1+ GSI 3D-VAR 3-km

15-min LHRRTM/

GoddardThompson

v3.4.1 None MYNN RUC 9-lev

Model Horiz/VertAdvection

ScalarAdvection

Upper-LevelDamping

6th Order Diffusion

SW RadiationUpdate Land Use MP Tend

LimitTime-Step

RAP 5th/5th Positive-Definite

w-Rayleigh0.2

Yes0.12 10 min MODIS

Fractional 0.01 K/s 60 s

HRRR 5th/5th Positive-Definite

w-Rayleigh0.2 No 5 min MODIS

Fractional 0.07 K/s 20-23 s

RAPv2 vs. HRRR – hourly updated

Model Run at: Domain Grid Points Grid Spacing Vertical Levels

Pressure Top

Boundary Conditions Initialized

RAP GSD,NCO

North America 758 x 567 13 km 50 10 mb GFS Hourly (cycled)

HRRR GSD, NCO Jul 2014 CONUS 1799 x

1059 3 km 50 20 mb RAP Hourly - RAP(no-cycle)

Courtesy of Stan Benjamin 24

13-km 6hr RAP reflectivity forecast HRRR 6hr reflectivity forecast

13-kmResolution

ParameterizedConvection

3-kmResolution

ExplicitConvection

HRRR (and RAP) Future MilestonesHRRR Milestones3-km HRRR – applications to aviation, severe weather, energy

ACCURATESTORM

STRUCTURE

ACCURATE ESTIMATE OFPERMABILITY

NOSTORM

STRUCTURE

NO ESTIMATE OF PERMEABILITY

07 June 2012

5 PM EDTRadar

observedCourtesy of Stan Benjamin 25

26

NCEP working with a team across NWS to make a set of 2-D fields on the 2.5km CONUS NDFD grid available in AWIPS and on NOAAPORT as close to Day 1 as possible

Requirements for the dataset are being driven by the field, but will be limited by the current capacity of the Satellite Broadcast Network (SBN). I suspect we'll have a phase 1, and then an expansion of fields after the upcoming SBN expansion

We will also put 2-D and 3-D output grids on the NCEP ftp server

Operational HRRR Data Distribution

27

• Model changes– 3-model system becomes 2-model system (NMMB and ARW)– Increase in both horizontal (16km to 12-14km) and vertical resolution

(35 to 45 levels)– Newer model version (3.4.1. to 3.4.2)

• Forecast diversity enhancement – Use three control analyses (GDAS, NDAS, RR)– Increasing physics diversity– Use of EnKF IC perturbations from hybrid Ens-VarGSI ensemble– Adding stochastic physics perturbation

• Post processing/calibration – Improve ensemble clustering algorithm (continuity in time)

• Ensemble products– Adding “Ensemble Anomaly Forecast” using climatology data – Adding more output fields tailored for renewable energy and dispersion

modeling communities

FY15 SREF Upgrade Attributes

GOALS

• Extend NEMS to be a flexible, unified, coupled operational framework that links NOAA modeling centers (EMC, GFDL, CPC, ESRL, SWPC)

• Initiate development of CFSv3 and improve predictive capability for medium range and longer time scales

• Promote interoperability and knowledge transfer across NOAA, DoD, NASA, NSF, and other agencies by using community interface standards

MODELING SYSTEM

• Models: NEMS GFS/GSM and NMM-B atmospheres+ HYCOM and MOM5 oceans + sea ice CICE model

• Infrastructure: ESMF-based, with additional conventions from the NUOPC consortium

• Timeline: Beginning scientific validation March 2014

MORE INFORMATION https://www.earthsystemcog.org/projects/nuopc/ncepplans

NOAA Environmental Modeling System (NEMS)Ocean and Ice Coupling

Courtesy of Cecelia DeLuca

29

Things That Keep Me Up at Night…

How will the suite evolve as the resolution of the global systems increase?• GFS capable of satisfying NAM requirements?• GEFS capable of satisfying SREF requirements?• Non-hydrostatic global model and medium range reforecasts from

GEFS becoming new requirements?

Will the regional systems shift to convection permitting applications?• HRRRE & NARRE capable of satisfying WOF requirements

What will the week 3+ to seasonal guidance system look like?• Can/should GEFS be extended to 30-days? (coupling required?)• What level of complexity is required in CFS?• Can the NMME be developed and sustained?

How does all the above impact down stream systems? How can NOAA engage the scientific community effectively? How can NOAA engage the stakeholders effectively?

NOAA Center for Weather and Climate Prediction

30

A.K.A.—the new building….• Four-story, 268,762 square foot

building in Riverdale, MD will house 800+ Federal employees, and contractors

• 5 NCEP Centers (NCO, EMC, HPC, OPC, CPC)

• NESDIS Center for Satellite Applications and Research (STAR)

• NESDIS Satellite Analysis Branch (SAB)• OAR Air Resources Laboratory

• Includes 465 seat auditorium & conference center, library, deli, fitness center and health unit

• Includes 40 spaces for visiting scientists

• Represents a “Game Changer” in our ability to do business