Storm-Scale Modeling with HRRR toward Warn-On-Forecast

Storm-Scale Modeling with HRRR toward Warn-On-Forecast
Technical Workshop on Numerical Guidance to Support WoF 5 February 2013 Storm-Scale Modeling with HRRR toward Warn-On-Forecast NOAA Earth System Research Laboratory GSD/AMB Stan Benjamin, Curtis Alexander, David Dowell, Steve Weygandt, Ming Hu, Tanya Smirnova, John Brown, Patrick Hofmann, Eric James, Ed Szoke, Haidao Lin

Gridpoint Statistical
RUC to Rapid Refresh transition CONUS domain RUC Model RUC 3dvar RUC post North American domain WRF-ARW Model GSI – Unipost Gridpoint Statistical Interpolation Hourly updated models 13km RUC 13km Rapid Refresh 3km HRRR NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

WRF-ARW / GSI enhancements for RAP
GSD contributions to WRF-ARW Grell 3-D cumulus scheme (updated each version) RUC (Smirnova) land-surface model (“ “ “) Diabatic Digital Filter Initialization (with NCAR) DFI-radar-latent heat reflectivity assimilation Modified MYNN PBL (v3.5) Latent heat limit to allow much longer time steps GSD contributions to GSI Non-variational cloud analysis Ingest for cloud / hydrometeor fields / obs Enhancements to surface obs assimilation / soil adjustment Options – aircraft V/T/RH reject list, temp bias correction NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

RUC  RAP / HRRR configuration
Community-based advanced model and analysis - WRF-ARW: advanced numerics, non-hydrostatic - GSI: advanced satellite data assimilation Model Run at: Domain Grid Points Grid Spacing Vertical Levels Pressure Top Boundary Conditions Initialized RUC GSD CONUS 451 x 337 13-km 50 Sigma/ Isentropic NAM Hourly (cycle) RAP NCO GSD North America 758 x 567 13 km 10 mb GFS (Part. cycle) HRRR 1799 x 1059 3 km 20 mb Hourly - RAP (no-cycle) Model Assimilation Cloud Analysis Radar DFI Radiation Microphysics Cum Param PBL LSM RUC RUC-3DVAR Yes RRTM / Dudhia Thompson (2003) Grell-Devenyi Burk –Thompson RUC 2003 RAP GSI-3DVAR RRTM/ Goddard G3 + Shallow MYJ 2010 HRRR None: RAP I.C No Dudhia None

Rapid Refresh Hourly Update Cycle
Hourly Observations RAP 2012 N. Amer Rawinsonde (T,V,RH) 120 Profiler – NOAA Network (V) 21 Profiler – 915 MHz (V, Tv) 25 Radar – VAD (V) 125 Radar reflectivity - CONUS 1km Radial wind Tested, late arrival Lightning (proxy reflectivity) NLDN, GLD360 Aircraft (V,T) 2-15K Aircraft - WVSS (RH) 0-800 Surface/METAR (T,Td,V,ps,cloud, vis, wx) Buoys/ships (V, ps) Mesonet (T, Td, V, ps) flagged GOES AMVs (V) AMSU/HIRS/MHS radiances Used GOES cloud-top press/temp 13km GPS – Precipitable water 260 WindSat scatterometer 2-10K Nacelle/Tower/Sodar 20/100/10 Partial cycle atmospheric fields – introduce GFS information 2x/day Cycle hydrometeors Fully cycle all land-sfc fields (soil temp, moisture, snow) 1-hr fcst Time (UTC) Analysis Fields 3DVAR Obs Back- ground

Radar reflectivity assimilation
Digital filter-based reflectivity assimilation initializes ongoing precipitation regions -20 min min Initial min min Backwards integration, no physics Forward integration,full physics with radar-based latent heating Initial fields with improved balance, storm-scale circulation RUC / RAP HRRR model forecast + RUC/RAP Convection suppression NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

Rapid Refresh (GSI + ARW) reflectivity assimilation example
Low-level Convergence Upper-level Divergence NSSL radar reflectivity (dBZ) 14z 22 Oct 2008 Z = 3 km K=4 U-comp. diff (radar - norad) K=17 U-comp. diff (radar - norad)

Adjust cycled explicit cloud fields using METAR and satellite data
Cloud and hydrometeor analysis Adjust cycled explicit cloud fields using METAR and satellite data YES HM Hydrometeor designation from radar NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

RAP soil adjustment from surface obs
Temp. dependent hydrometeor specification RAP soil adjustment from surface obs From radar 6h RAP fcst dewpoint bias Eastern US (Fcst - obs) Specify hydrometeors Specify only snow hydrometeors YES HM K WARM (Tsfc > 5C) COLD (Tsfc < 5C) Soil adjust (T’soil , Q’soil ) Q’soil applied if: -- T’(k=1) q’(k=1) opp. sign -- Daytime, No clouds -- Proportional to q’(k=1) Assume – Bowen ratio error from soil Q’ error Apply top 2 LSM levels Moist Bias Dry Bias High afternoon dewpoint bias NO soil adjust WITH soil adjust

pseudo-observations (from sfc Td obs)
Assimilate PBL-based pseudo-observations (from sfc Td obs)  Create moisture pseudo-observations through depth of mixed BL Greater retention of surface dewpoint information

RAPv1 RAPv2 March 2, 2012: OH/TN Valley Tornado Outbreak
1-hr forecast SBCAPE Valid 21Z RAPv1 RAPv2 RAP-v2 instability forecasts better than RAP

Hourly HRRR Initialization from RAP
13 km RAP Obs Obs Obs GSI 3D-VAR GSI 3D-VAR GSI 3D-VAR HM Obs HM Obs HM Obs Cloud Anx Cloud Anx Cloud Anx 1 hr fcst 1 hr fcst Refl Obs Refl Obs Refl Obs Digital Filter Digital Filter Digital Filter 18 hr fcst 18 hr fcst 18 hr fcst 3 km HRRR 3-km Interp 3-km Interp 3-km Interp 15 hr fcst 15 hr fcst 15 hr fcst

13km Rapid Refresh and 3km HRRR hourly updated weather models
ESRL – experimental version NWS-NCEP - operational RAPv1 – used in 2011 Initialized 2011 HRRR effective but too many storms RAPv2 – used in 2012 Initialized 2012 HRRR Better use of surface obs / radar, storm bias eliminated RAP-2013 – 3km/15min radar assim, GOES cloud assim, ensemble assim HRRR – 2012 Major improvement over 2011 HRRR, storm coverage/accuracy HRRR – 2013 3km/15min radar assimilation Initialized from RAP-2013 Available 45 min earlier, much more accurate 0-15h storm forecasts, more reliable 2-computer Implemented 1 May 2012 RAPv2 - Scheduled to be implemented in spring 2014 HRRR – estimated 2015

RAP and HRRR Config 2011 NCEP Production Suite Review
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 CONUS 1799 x 1059 3 km 20 mb Hourly - RAP (no-cycle) Model Version Assimilation Radar DFI Radiation Microphysics Cum Param PBL LSM RAP WRF-ARW v3.2.1+ GSI-3DVAR Yes RRTM/Goddard Thompson G3 + Shallow MYJ RUC HRRR WRF-ARW V3.2.1+ None: RAP I.C. No RRTM/ Dudhia None Model Horiz/Vert Advection Scalar Upper-Level Damping SW Radiation Update Land Use MP Tend Limit Time-Step RAP 5th/3rd Monotonic Diffusive 0.02 30 min USGS 0.01 K/s 60 s HRRR 0.10 K/s 18-23 s NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

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 CONUS 1799 x 1059 3 km 20 mb Hourly - RAP (no-cycle) Model Version Assimilation Radar DFI Radiation Microphysics Cum Param PBL LSM RAP WRF-ARW v3.3.1+ GSI-3DVAR Yes RRTM/Goddard Thompson G3 + Shallow MYJ RUC HRRR WRF-ARW v3.3.1+ None: RAP I.C. No RRTM/Dudhia None Model Horiz/Vert Advection Scalar Upper-Level Damping SW Radiation Update Land Use MP Tend Limit Time-Step RAP 5th/5th Positive-Definite w-Rayleigh 0.2 10 min MODIS Fractional 0.01 K/s 60 s HRRR 5 min 0.07 K/s 20-23 s NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

HRRR June Retro Verification
Reflectivity Eastern US 160 Runs 29 May – 12 June 2011 Reflectivity ≥ 30 dBZ HRRR 2011 HRRR 2012 40 km CSI 03 km BIAS Optimal Reduced high bias in first 6 hrs and improved CSI NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

HRRR Retro Case Studies
HRRR 5hr fcst 2011 Real-Time 21z 01 June 2011 Observations HRRR 5hr fcst 2012 Version Composite Reflectivity (dBZ) Large reduction in false alarm (excessive) convection Improved structure to broken convective line NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

HRRR Retro Case Studies
HRRR 10hr fcst 2011 Real-Time 18z 07 June 2011 Observations HRRR 10hr fcst 2012 Version Composite Reflectivity (dBZ) Improved mesoscale convective system (MCS) maintenance NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

HRRR Retro Case Studies Reduction in false alarm convective initiation
HRRR 8hr fcst 2011 Real-Time 00z 31 May 2011 Observations HRRR 8hr fcst 2012 Version Composite Reflectivity (dBZ) Reduction in false alarm convective initiation Retro Assessment High-Resolution Rapid Refresh 21 June 2012

NOAA High-Resolution Rapid Refresh HRRR Model
Forecast of Mid-Atlantic Derecho – 29 June 2012 Radar observed HRRR forecast Composite Reflectivity (dBZ)

(Would need lengthy RAP retro to spin-up soil fields)
Radar Observed RAP version 1 RAP version 2 [2012 GSD real-time run] post-DFI [2011 RAP/HRRR code] (Would need lengthy RAP retro to spin-up soil fields) 03z 30 June 12h fcsts RAP version 1 RAP version 2 [2012 NCEP real-time run] [2012 GSD real-time run] pre-DFI pre-DFI * Only “pre-DFI” fields available from NCEP RAP

RAP version 2 0 h fcsts RAP version 1 RAP version 2
Radar Observed RAP version 2 Init post-DFI 15z 29 June 0 h fcsts RAP version 1 RAP version 2 Init pre-DFI Init pre-DFI

Radar Observed RAP version 2 Init post-DFI 00z 30 June 9 h fcsts RAP version 1 RAP version 2 Init pre-DFI Init pre-DFI ~ 11z LTG

RAP version 2 12h fcsts RAP version 1 RAP version 2
Radar Observed RAP version 2 Init post-DFI 03z 30 June 12h fcsts RAP version 1 RAP version 2 Init pre-DFI Init pre-DFI Conclusions: RAPv2 DA for mesoscale environment critical. ~ 11z LTG

NOAA Next-Generation HRRR System
Forecast of 2011 Alabama Tornado Outbreak Tuscaloosa tornadic supercell 5 PM CDT 27 April 2011 HRRR storms 9-h forecast Radar observed storms Reflectivity (dBZ) Forecast available by 10 AM CDT

RAP version 2 Changes RAP Data Assimilation Model (13 km)
GFS ensemble background error covariance Merge with recent GSI trunk Sfc. moisture pseudo-obs Sfc. obs-based soil adjust Temp. dependent radar hydrometeor building + clearing Cloud assimilation mods Cloud fraction assimilation Full column cloud building Radiance bias correction Aircraft temp bias correction NLDN/GLD360 lightning assimilation (proxy reflectivity) WRFv incl. physics changes (convection, snow-radiation fix) Numerics changes 6th-order diffusion near surface 5th-order vertical advection w-damp upper bound conditions Physics changes MYNN PBL 9-layer RUC LSM (from 6-layer) MODIS land-use (fractional) Modified roughness length Revised Thompson cloud microphysics RRTMG shortwave/longwave radiation GFS ensemble background error covariance

3h fcst – 2m temp bias vs. METARs – e. US 3h fcst - 10m wind bias vs. METARs – e. US RAPdev1 shows more accurate 2m temp diurnal bias, less exaggerated diurnal cycle than evidence in RAP-primary (RR1h) RAPdev1 shows less 10m wind bias compared to RAP-primary, near 0 in daytime, from 0.8 m/s down to ~0.5 at night (an improvement) RUC LSM with 9 levels – implemented in RAP-dev1 on 18 Oct12: Zs= 0, 1, 4, 10, 20, 40, 100, 160, 300 cm RUC LSM with 6 levels –in RAP primary ESRL, RAP-oper-NCEP: Zs= 0, 5, 20, 40, 160, 300 cm

RAP-2012 – using GSI using fixed background error covariance
Obs Obs Obs GSI 3D-Var GSI 3D-Var GSI 3D-Var HM Obs HM Obs HM Obs Cloud Anx Cloud Anx 1 hr fcst Cloud Anx Copy 6-h and 9-h GFS EnKF ensemble forecast member (80) four times a day GSI search the nearest ensemble forecast and do hybrid Ensemble forecast valid time is different from analysis time 1 hr fcst Refl Obs Refl Obs Refl Obs Digital Filter Digital Filter Digital Filter 18 hr fcst 18 hr fcst 18 hr fcst 13 km RAP 13z 14z 15z

RAP-2013 GSI hybrid using Global Ensemble
Available 4 times a day valid at 03, 09, 15, 21Z 80-member GFS EnKF Ensemble forecast valid at 15Z (9-h fcst from 6Z) Obs Obs Obs GSI Hybrid GSI Hybrid GSI Hybrid HM Obs HM Obs HM Obs Cloud Anx Cloud Anx 1 hr fcst Cloud Anx Copy 6-h and 9-h GFS EnKF ensemble forecast member (80) four times a day GSI search the nearest ensemble forecast and do hybrid Ensemble forecast valid time is different from analysis time 1 hr fcst Refl Obs Refl Obs Refl Obs Digital Filter Digital Filter Digital Filter 18 hr fcst 18 hr fcst 18 hr fcst 13 km RAP 13z 14z 15z

RMS profile for 3h forecasts
Real-time test for RAP hybrid using bkg error cov from GDAS hybrid DA ensemble Compare RAP development with GSI hybrid to RAP primary cycle with GSI-3dvar 30-day real-time test from 22 Nov – 22 Dec 2012 GSI hybrid with half static BE and half BE from GFS ensemble (80 member) forecasts RMS profile for 3h forecasts RAP hybrid RAP

RAP and HRRR Changes 2013 Model Data Assimilation RAP-ESRL (13 km)
WRFv incl. physics changes (convection, snow-radiation fix) Numerics changes: 6th-order diffusion near surface Physics changes: MYNN PBL scheme 9-layer RUC LSM (from 6-layer) Modified roughness length RRTMG short/longwave radiation Thompson microphysics update Merge with GFS trunk GFS ensemble background error cov Cloud fraction assimilation Full column cloud building Radar hydrometeor build/clear Improved cloud hydrometeor analysis Aircraft temp bias correction Radiance bias correction Reduced observation error (sharper inversions, low-level thermo) HRRR (3 km) 3 km/15 min reflect assimilation 3 km cloud cycling 3 km land-surface cycling

HRRR Real-Time Exper Products
Background Radar Specification of Hydrometeors Scale at which Latent Heating is applied Dimensionality Updated 2012 HRRR model initialization 13-km RAP No 13-km 3-D Hourly 2013 HRRR yes 3km in 60min spin-up (also using 13km radar-LH-DFI) Rapidly Updating Analysis (RUA-HRRR) 3-km HRRR 1 hr fcst Yes None Real-Time Meso Analysis (RTMA-HRRR) 2-D (15 min planned) Time-Lagged HRRR (HCPF) Fcsts Same as HRRR RUA Meeting High-Resolution Rapid Refresh 25 Jan 2013

ESRL GSI applications at 3km 1800 x 1060 (x 50 levs)
CPU cores (zeus) Run time 3km HRRR (3d, radar, sat only, no var solver) 240 8 min 3km RUA (no var) 84 6-7 min 3km RTMA (2d, sfc obs only) 60 7 min Full 3km 3dvar (full RAP obs data set) 300 20-30 min

2013: Cycled Reflectivity at 3 km
13 km RAP 13z 14z 15z Obs Obs Obs GSI 3D-VAR GSI 3D-VAR GSI 3D-VAR HM Obs HM Obs HM Obs Cloud Anx Cloud Anx Cloud Anx 1 hr fcst 1 hr fcst Refl Obs Refl Obs Refl Obs Digital Filter Digital Filter Digital Filter 18 hr fcst 18 hr fcst 18 hr fcst 3 km HRRR 3-km Interp 1 hr pre-fcst 15 hr fcst 1-hr Reduction In Latency for 14z HRRR 3-km Interp Refl Obs

Latent Heating (LH) Specification
Temperature Tendency (i.e. LH) = f(Observed Reflectivity) LH specified from reflectivity obs applied in four 15-min periods The observations are valid at the end of each 15-min pre-fcst period NO digital filtering at 3-km Hour old mesoscale obs Latency reduced by 1 hr -60 -45 -30 -15 LH = Latent Heating Rate (K/s) p = Pressure Lv = Latent heat of vaporization Lf = Latent heat of fusion Rd = Dry gas constant cp = Specific heat of dry air at constant p f[Ze] = Reflectivity factor converted to rain/snow condensate t = Time period of condensate formation (600s i.e. 10 min) Observed 3-D Radar Reflectivity Time (min) Model Pre-Forecast Time (min) RUA Meeting High-Resolution Rapid Refresh 25 Jan 2013

Latent Heating (LH) Specification
Many options for weighting LH specification vs model LH Two approaches including: 100% specification for the entire pre-forecast hour (b) Time-varying with linear ramp down to 0% specification at 1 hr Option (b) permits more “free model” behavior before additional DA Experiment 1a “Fixed” Experiment 1b “Ramp” 1 1 Observed Radar Reflectivity (3-D) Based Latent Heating Weight Weight Model Microphysics Latent Heating Model Pre-Forecast Time (min) Model Pre-Forecast Time (min) RUA Meeting High-Resolution Rapid Refresh 25 Jan 2013

Benefit from pre-forecast 3-km model integration
Obs 0000 UTC 11 June 2011 0-h fcst without 3-km radar DA 0-h fcst with 3-km radar DA (fixed) 0-h fcst with 3-km radar DA (ramp) Benefit from pre-forecast 3-km model integration

Convective systems more mature even by 1-hr
Obs 0100 UTC 11 June 2011 1-h fcst without 3-km radar DA 1-h fcst with 3-km radar DA (fixed) 1-h fcst with 3-km radar DA (ramp) Convective systems more mature even by 1-hr

HRRR Reflectivity Verification
3-day retrospective period June 2011 (36 runs) Forecasts every 2 hours > 25 dBZ Composite Reflectivity Eastern half of US With 3-km fixed radar DA With 3-km ramp radar DA Without 3-km radar DA Bias = 1.0 Upscaled to 40-km grid Native 3-km grid Greatly improved CSI and BIAS between 0-1 fcst hr Benefit persists until 4 hrs Very similar skill at longer lead times

HRRR Reflectivity Verification
14-day retrospective period June 2011 (160 runs) Forecasts every 2 hours > 25 dBZ Composite Reflectivity Eastern half of US With 3-km ramp radar DA Without 3-km radar DA Bias = 1.0 Upscaled to 40-km grid Native 3-km grid Greatly improved CSI and BIAS between 0-1 fcst hr Benefit persists until 4 hrs Very similar skill at longer lead times

Precip hydrometeor (qr/qs/qg) clearing/building from 3-D radar reflectivity data
NCEP RAPv1 ESRL RAPv2 (primary) until 29 Jan 2013 ESRL RAPv2 (primary) as of 06 UTC 29 Jan 13 Tsfc ≤ 5C Tsfc > 5C building All levels with sufficiently high obs refl (> 0 dBZ) All levels (same as for Tsfc < 5C) None Specify qs at all levels with sufficiently high reflectivity (> 0 dBZ) Set to max diagnosed qs only at that obs max refl level. clearing Clear except allow diagnosed qs to be retained at max obs refl level Clear qr/qs/qg in any volumes with obs refl < 0 dBZ

2m temps below are clearly < 4 deg C,
In fact, usually < -10 deg C

Observed lake effect snow band is clearly over Lake Ontario

But the lake effect snow band in the HRRR 0h is located over land (NY state) south of Lake Ontario

With snow building/clearing Without snow building/clearing
Improved snowfall/snowcover 1- hr Forecasts valid 14 UTC 25 Jan 2013 With snow building/clearing Without snow building/clearing

Improved 0-hr reflectivity analysis
Introduced snow building/clearing in ESRL RAPv2 (primary) and implicitly ESRL HRRR (primary and dev1) At 06 UTC 29 January 2013 0-hr 13-km > 25 dBZ 0-hr 20-km > 25 dBZ

Improved reflectivity forecasts
hr forecasts for HRRR-primary and ESRL-RAP-primary are now performing fairly well 2. Addition of 3km/15min assimilation (HRRR_dev1) helps further still at 0-2 hrs 20-km > 30 dBZ 20-km > 35 dBZ

HRRR Real-Time Exper Products
Background Radar Specification of Hydrometeors Scale at which Latent Heating is applied Dimensionality Updated 2012 HRRR model initialization 13-km RAP No 13-km 3-D Hourly 2013 HRRR yes 3km in 60min spin-up (also using 13km radar-LH-DFI) Rapidly Updating Analysis (RUA-HRRR) 3-km HRRR 1 hr fcst Yes None Real-Time Meso Analysis (RTMA-HRRR) 2-D (15 min planned) Time-Lagged HRRR (HCPF) Fcsts Same as HRRR RUA Meeting High-Resolution Rapid Refresh 25 Jan 2013

Hourly HRRR RTMA 13z 14z 15z 3 km HRRR Obs Obs GSI 2D-VAR GSI 2D-VAR
1 hr fcst 1 hr fcst 3-km Interp 3-km Interp 3-km Interp 15 hr fcst 15 hr fcst 15 hr fcst

(Background) Valid 19 UTC
3-km HRRR RTMA 1-hr HRRR Fcst (Background) Valid 19 UTC 30 Nov 2012 0-hr HRRR Anal Valid 19 UTC 30 Nov 2012 Analysis Increments RTMA HRRR Anx 10 m Winds RUA Meeting High-Resolution Rapid Refresh 25 Jan 2013

Hourly HRRR Rapidly Updating Anx
13z 14z 15z 3 km HRRR HM Obs HM Obs Cloud Anx Cloud Anx 1 hr fcst 1 hr fcst 3-km Interp 3-km Interp 3-km Interp 15 hr fcst 15 hr fcst 15 hr fcst

3-km HRRR RUA Specifies Rapidly Hydrometeors Updating From Radar
Obs 22 UTC 03 Nov 2012 Specifies Hydrometeors From Radar Observations Rapidly Updating Analysis (RUA) Cloud Anx GSI 1-hr HRRR Forecast (Background) Valid 22 UTC 03 November 2012 0-hr HRRR Analysis (RUA) Valid 22 UTC 03 November 2012 RUA Meeting High-Resolution Rapid Refresh 25 Jan 2013

Time-Lagged (TL) Ensemble
Forecasts valid 21-22z 27 April 2011 Forecasts valid 22-23z 27 April 2011 10-11 hr fcst 11-12 hr fcst HRRR 11z Init 09-10 hr fcst 10-11 hr fcst HRRR 12z Init 08-09 hr fcst 09-10 hr fcst HRRR 13z Init All six forecasts combined to form probabilities valid 22z 27 April 2011 Spatial radius 45 km Time radius 1 hr UH threshold 25 m2/s2

TL Example: 22 May 2011 Joplin, MO AMS 26th SLS Conf
13z + 11hr fcst Valid 00z 23 May 2011 1300z SPC Tornado Probability Joplin, MO 22 May 2011 Storm Reports Tornadic Storm Probability (%) Tornado = Red Dots AMS 26th SLS Conf High-Resolution Rapid Refresh 06 Nov 2012

TL Example: 22 May 2011 Joplin, MO AMS 26th SLS Conf
13z + 11hr fcst Valid 00z 23 May 2011 Observed Reflectivity 00z 23 May Joplin, MO 22 May 2011 Storm Reports Tornadic Storm Probability (%) Reflectivity (dBZ) Tornado = Red Dots AMS 26th SLS Conf High-Resolution Rapid Refresh 06 Nov 2012

HRRR Reliability HRRR (and RAP) Future Milestones HRRR Milestones
HRRR 12 hr fcst availability Includes all missed/incomplete runs Jet Maintenance Jet (HRRR primary) Zeus (HRRR backup) HPCRAC High-Resolution Rapid Refresh 22 Jan 2013

HRRR Reliability HRRR (and RAP) Future Milestones HRRR Milestones
HRRR 12 hr fcst availability Includes all missed/incomplete runs Jet or Zeus (union) HPCRAC High-Resolution Rapid Refresh 22 Jan 2013

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 CONUS 1799 x 1059 3 km 20 mb Hourly - RAP (no-cycle) Model Version Assimilation Radar DFI Radiation Microphysics Cum Param PBL LSM RAP WRF-ARW v3.3.1+ GSI-3DVAR Yes RRTM/Goddard Thompson G3 + Shallow MYJ RUC HRRR WRF-ARW v3.3.1+ None: RAP I.C. No RRTM/Dudhia None Model Horiz/Vert Advection Scalar Upper-Level Damping SW Radiation Update Land Use MP Tend Limit Time-Step RAP 5th/5th Positive-Definite w-Rayleigh 0.2 10 min MODIS Fractional 0.01 K/s 60 s HRRR 5 min 0.07 K/s 20-23 s NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

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 CONUS 1799 x 1059 3 km 20 mb Hourly - RAP (no-cycle) Model Version Assimilation Radar DFI Radiation Microphysics Cum Param PBL LSM RAP WRF-ARW v3.4.1+ GSI-hybrid Yes RRTMG/ LW/SW Thompson v3.4 G3 v3.5 + Shallow MYNN/MYJ? RUC 9-lev HRRR WRF-ARW V3.4.1+ RAP I.C. +3km/15min 3km LH None Model Horiz/Vert Advection Scalar Upper-Level Damping SW Radiation Update Land Use MP Tend Limit Time-Step RAP 5th/5th Positive-Definite w-Rayleigh 0.2 10 min MODIS Fractional 0.01 K/s 60 s HRRR 5 min 0.07 K/s 20-23 s NCEP Production Suite Review Rapid Refresh / HRRR 4-5 December 2012

13km Rapid Refresh and 3km HRRR hourly updated weather models
ESRL – experimental version NWS-NCEP - operational RAPv1 – used in 2011 Initialized 2011 HRRR effective but too many storms RAPv2 – used in 2012 Initialized 2012 HRRR Better use of surface obs / radar, storm bias eliminated RAP-2013 – 3km/15min radar assim, GOES cloud assim, ensemble assim HRRR – 2012 Major improvement over 2011 HRRR, storm coverage/accuracy HRRR – 2013 3km/15min radar assimilation Initialized from RAP-2013 Available 45 min earlier, much more accurate 0-15h storm forecasts, more reliable 2-computer Implemented 1 May 2012 RAPv2 - Scheduled to be implemented in spring 2014 HRRR – estimated 2015

North American Rapid Refresh ENSEMBLE (NARRE)
? North American Rapid Refresh ENSEMBLE (NARRE) NMMB (from NCEP) & ARW (from ESRL) dynamic cores Common NAM parent domain at km Initially ~6 member ensemble made up of equal numbers of NMMB- & ARW-based configurations Hourly updated with forecasts to 24 hours EnKF/hybrid data assimilation for ensemble member initialization Use 3 different physics suites for ensemble members but not for data assimilation RAP physics (Thompson microphysics, Grell-3d cu, MYJ or MYNN PBL) NMM physics (Ferrier microphysics, BMJ/Janjic cu, MYJ PBL) NCAR physics (let MMM recommend - WSM,YSU,KF?) Consider stochastic physics for a single physics suite Sept 2012 – modifications from Stan B. and Geoff D.

NARRE hyb/EnKF assim options – p.1
Separate 1h ensemble data assimilation cycle Testing already performed by CAPS – Ming Xue, Kefeng Zhu – 40km/3h RAP EnKF assimilation GSD – Ming Hu – 13km/40km/1h RAP EnKF assimilation Use of GFS (global) EnKF-produced background error covariance Only updated every 6h, not available at 1h freq Current promising testing by Dave Parrish and Wan-shu Wu at NCEP in 6h NAM cycle. ESRL (Ming Hu) with alternative RAP now running since Nov 2012 Clearly outperforms use of fixed background error covariance (current NCEP NAM and NCEP RAP)

NARRE hybrid/EnKF assimilation options – p2
Separate 1h ensemble data assimilation cycle 12-40km resolution, 80 members 80 (members) / 27 - ~3h 12km forecast 27 – 3x resolution **3 – 36km Ensemble Use of GFS (global) EnKF-produced background error covariance Possible interim solution for 1h cycle – require hourly output from GFS hybrid from 3h-9h, rescale for forecast duration for NARRE Other limitations Resolution Future extension to cloud/hydrometeor and radar assimilation components. Anticipated skill Fixed BEC (current) < GFS EnKF BEC < NARRE 1h BEC Jan 13 update – RAP testing using GSI-hybrid DA with GFS-DA-ensemble fully successful, will be promoted to primary ESRL RAP (initializing HRRR) in February 2013

Assuming 1h EnKF assimilation for NARRE
Resolution for data assimilation ensemble 30-40km may be sufficient (and perhaps higher-res cannot be afforded) Model for data assimilation EnKF Choose single dynamic core (ARW or NMMB), do not mix cores Physics for data assimilation EnKF Apply stochastic physics to a single set Successful inflation/localization in RAP ensemble-RAP testing by CAPS, NSSL, U.OK

High Resolution Rapid Refresh ENSEMBLE (HRRRE)
? High 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 (EnKF+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 HPCC funding over and above that required for the NARRE From Geoff DiMego, Dec 2011, NCEP Model Review

DRAFT Storm Prediction Center Desired Numerical Guidance Attributes
74

DRAFT Storm Prediction Center Desired Numerical Guidance Attributes
Possible alternative NWP 2-member suite Global ensemble – as shown by EMC  10-12km global member ensemble runs to Day 14 twice daily (to 7d at 06z, 18z) hourly assimilation cycle with 24h runs init hourly Single regional model/assimilation ensemble - SSEF SSEF (equivalent to HRRRe) – 2-3km Initialize hourly (or subhourly), run to 24h Every 6h, run out to 48-60h 2-3km EnKF/hybrid/other DA Multi-species microphysics with at least 2-moment rain Include aerosols/fire/smoke for all runs with aerosol-aware microphysics No separate HWRF or fire-weather runs or AQ/dust run km multi-WFO nest where needed (WoF) SSEF analysis = RUA, no separate products 75

Global  Continental  Local Nesting Strategy
Discussion items Nesting necessary to save computing; current NAM domain could be expanded to global 10 km Advantages/disadvantages of same model for global and regional Applicability (DA) of a multi-model ensemble-based covariance estimate with multiple bias characteristics Data assimilation at resolution of 3 km or less poses challenges Advanced assimilation of radar reflectivity Consistent integration of radar and satellite info Proper dynamically balanced analysis increments at multiple high resolution scales (10, 3, 1 km)

DCMIP goal: Identify global model numerical deficiencies, especially for inadvertent or intentional diffusion

March 2, 2012 Major Tornado Outbreak – Ohio and Tennessee Valleys
9 killer tornadoes (2 EF-4; 4 EF-3; 2 EF-2, 1 EF-1) 40 fatalities, numerous injuries and widespread damage Tornado Damage in Henryville, IN SPC report – NCEP model review – Dec12 – Weiss, Jirak, etc.

SPC report – NCEP model review – Dec12 – Weiss, Jirak, etc.
Simulated Reflectivity (1 km AGL) hr Forecasts Valid 00z 3 March 2012 NAM 4 km CONUS Nest NAM 1.33 km Fire Nest Both NAM nests develop solid line srn KY/TN with higher reflectivity in Fire Nest NSSL ARW better resolves cellular structures including ahead of QLCS NSSL 4 km WRF-ARW Observed Radar 2 km Mosaic

Zoom-in shows “effective resolution” of NAM can be less than NSSL-ARW
SPC report – NCEP model review – Dec12 – Weiss, Jirak, etc. Simulated Reflectivity (1 km AGL) hr Forecasts Valid 00z 3 March 2012 NAM 4 km CONUS Nest NAM 1.33 km Fire Nest Zoom-in shows “effective resolution” of NAM can be less than NSSL-ARW NSSL 4 km WRF-ARW Observed Radar 2 km Mosaic

Option: MPAS as a unified global to storm-scale model for NOAA applications
High-order accurate numerics Has performd well in DCMIP idealized tests Current nest via polygon tesselation Need to add 1-way separate nest with LBC from global model but this can be done. Would also provide unified weather-seasonal-climate model Community support (although not intended by NCAR to be as one-fits-all as WRF)

From HRRR/HRRRe to WoF Radar-DFI-LH effective start at 13km/3km
Mesoscale environment via GSI enhancements for RAPv2+ critical for HRRR 3km/80-member hourly updated GSI-hybrid data assimilation over HRRR-CONUS domain plausible Estimated needed 10,000 cores (Jeff W. / Stan) Model design for HRRR – most accurate numerics possible necessary for identification of storm structure (line vs. supercells vs. cell vs. MCS variations) ARW, no 6th order diffusion, 5th-order vertical advection WoF options – discussion this week with NSSL, ESRL, EMC, U.OK, SPC partners direct nest inside HRRR/HRRRe within same executable Separate system using HRRR/HRRRe lateral BCs/background

Storm-Scale Modeling with HRRR toward Warn-On-Forecast

Similar presentations

Presentation on theme: "Storm-Scale Modeling with HRRR toward Warn-On-Forecast"— Presentation transcript:

Similar presentations

About project

Feedback

Log in

Auth with social network:

Storm-Scale Modeling with HRRR toward Warn-On-Forecast

Similar presentations

Presentation on theme: "Storm-Scale Modeling with HRRR toward Warn-On-Forecast"— Presentation transcript:

Similar presentations

About project

Feedback