1 / 29

SPC Evaluation of the NAM-WRF Model: Spring Experiment 2006 NOAA Hazardous Weather Testbed

SPC Evaluation of the NAM-WRF Model: Spring Experiment 2006 NOAA Hazardous Weather Testbed. Steven Weiss and Jack Kain May 2006. Spring Experiment 2006. The 2006 Experiment consisted of a pre-implementation evaluation of the 12 km WRF-NMM Examined 12z run (Day 1) and 00z runs (Day 2 and Day3)

scott
Télécharger la présentation

SPC Evaluation of the NAM-WRF Model: Spring Experiment 2006 NOAA Hazardous Weather Testbed

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. SPC Evaluation of the NAM-WRF Model:Spring Experiment 2006NOAA Hazardous Weather Testbed Steven Weiss and Jack Kain May 2006

  2. Spring Experiment 2006 • The 2006 Experiment consisted of a pre-implementation evaluation of the 12 km WRF-NMM • Examined 12z run (Day 1) and 00z runs (Day 2 and Day3) • SPC focused on the performance of the NAM-WRF and NAM-Eta during 16 spring severe weather episodes • Mostly post-event assessments • Comparative performance between NAM-WRF and NAM-Eta from SPC forecaster perspective • Implications of using NAM-WRF as primary regional model input for SPC severe weather outlooks

  3. Model Evaluation Methodologies • Looked at 8 key model fields for Outlook-scale severe weather forecasting plus soundings • 250 mb heights/isotachs (upper level jet) • 500 height/vorticity (mid level pattern/features) • 850 mb height/isotachs (low level jet) • EMSL (surface pattern/boundaries) • 2m dew point (moisture) • MUCAPE (instability) • 0-6 km vector shear (deep layer shear) • 3-hr accumulated precipitation • Although all fields examined, findings will focus on moisture, sounding structure, and precipitation

  4. Model Evaluation Methodologies • SPC created special web pages for: • Display of four-panel graphics of NAM-WRF and NAM-Eta forecasts, difference fields, and “verifying” RUC or SPC sfcoa analyses • Time-matched from 00-84 hrs • Comparison of model soundings and observed soundings, including vertical difference fields • At radiosonde locations from 00-60 hrs • Input of evaluation findings, including subjective evaluation ratings of forecast skill • See www.spc.noaa.gov/exper/Spring_2006/

  5. NAM-WRF Key Findings - Moisture • 2m dew point generally improved NAM-Eta sfcoa • 00hr initial moisture fields at 00z/16 Apr • Eta too moist in warm sector • WRF drier than Eta and closer to observed data at initial time NAM-WRF (WRF minus Eta)

  6. NAM-WRF Key Findings - Moisture • 2m dew point generally improved NAM-Eta sfcoa • 84 hr forecast of return flow moisture valid 00z/5 Apr • WRF better with leading edge of 50s dew points staying south of Red River NAM-WRF (WRF minus Eta)

  7. NAM-WRF Key Findings - Moisture • 2m dew point generally improved NAM-Eta sfcoa • 18 hr forecast of warm sector moisture valid 06z/8 Apr • WRF better with moisture in warm sector over the Southeast; Eta predicts values too high NAM-WRF (WRF minus Eta)

  8. NAM-WRF Key Findings - Moisture • However, on rare occasion spurious drying in TX NAM-Eta sfcoa • 54 hr forecast of warm sector moisture valid 06z/8 Apr • WRF shows dry region in TX/OK • Drying appeared to impact CAPE and precipitation forecasts during 2 day period • Evident in forecast soundings for this episode NAM-WRF (WRF minus Eta)

  9. 60 hr WRF and Eta Soundings at FWD valid 12z 8 Apr • WRF (red/green) considerably drier in low levels compared to Eta (purple) • Also note “noisy” vertical structure in WRF low levels above PBL compared to much smoother Eta profiles • This characteristic was observed in WRF soundings on a number of occasions

  10. 24hr WRF and Observed Soundings at CRPvalid 00z 4 Apr • WRF (red/green) exhibits considerably shallower PBL depth compared to CRP raob (purple) • The shallow PBL structure in the WRF was noted many times with southeast low level winds at CRP • Is this related to PBL structure over adjacent Gulf of Mexico?

  11. NAM-WRF Key Findings - Soundings • Forecast Soundings • SPC forecasters examine soundings frequently for temperature and moisture information in pre-convective environment • PBL structure most important (depth of moist layer, capping inversion, etc.) • Neither the WRF nor the Eta consistently handled PBL evolution reliably • This is a continuing challenge

  12. NAM-WRF Evaluation - Precipitation • Precipitation Forecasts • 3-hr accumulated precipitation examined for timing and location for model convective development • Focus on mesoscale evolution of model precipitation and correspondence with observed radar or multi-sensor precipitation • Initiation • Direction and speed of system movement • Areal coverage • Configuration and orientation of mesoscale features • Perceived convective mode

  13. Example 1 – 2 April 2006 • Strongly forced, very active severe day • Killer tornadoes in IL, MO and TN causing 26 deaths

  14. NAM-WRF Evaluation - Precipitation • Sometimes WRF precipitation was better NAM-Eta Observed Pcpn • Day 1 Forecast of15 hr precipitation valid 03z/3Apr • Eta much delayed in developing severe storms IL into KY, TN and AR • WRF better than Eta (but also slow with initial severe storms) NAM-WRF (WRF minus Eta)

  15. NAM-WRF Evaluation - Precipitation • Sometimes WRF precipitation was worse NAM-Eta Observed Pcpn • Day 3 Forecast of 75 hr precipitation valid 03z/3Apr • Eta too far west with severe storms IL into KY/TN and AR • This WRF run failed to develop storms in TN Valley until 6 hrs after this time NAM-WRF (WRF minus Eta)

  16. Example 2 – 2 May 2006 • Weakly forced regional severe day • Tornadoes, hail larger than 4” and 85 mph winds

  17. NAM-WRF Evaluation - Precipitation • Sometimes neither model was good NAM-Eta Observed Pcpn • Day 2 Forecast of54 hr precipitation valid 06z/3May • Neither model predicted night time TX severe storms well NAM-WRF (WRF minus Eta)

  18. Example 3 – 29 April • Period of low synoptic scale predictability • SREF showed large spread • Local severe weather day • Tornadoes and wind damage southeast TX and southern LA

  19. NAM-WRF Evaluation • SREF Spaghetti chart of 5460 m 500 mb height contour • Enhanced uncertainty in upper low evolution over Central States

  20. NAM-WRF Evaluation • Large 500 mb differences between WRF and Eta NAM-Eta RUC Analysis • 72 hr Eta forecast valid 00z 30 Apr correctly predicted single upper low • WRF forecast shows double upper low structure in Central US • Evolution of boundaries, moisture, and instability reflected differing synoptic patterns NAM-WRF (WRF minus Eta)

  21. NAM-WRF Evaluation - Precipitation • During late April case the Eta pattern was better NAM-Eta Observed Pcpn • Day 3 Forecast of 72 hr precipitation valid 00z/30 Apr • More accurate Eta synoptic scale forecast resulted in better prediction of storms NAM-WRF (WRF minus Eta)

  22. Subjective Evaluation Scores • Subjective ratings of 0 to 10 were assigned for Day 1, Day 2, and Day 3 forecast periods for 16 active severe weather days • 0 rating – very poor correspondence with observed wx • Poor precipitation guidance to severe weather forecasters • 10 rating – excellent correspondence with observed wx • Excellent precipitation guidance to severe weather forecasters

  23. Mean Subjective Precipitation Scores • WRF evaluation scores better on Day 1 • Eta evaluation scores higher on Day 2 and Day 3 • Caveat – small sample size

  24. NAM-WRF Evaluation - Reflectivity • Reflectivity now available from 12 km NAM-WRF • Total reflectivity contributions from grid scale microphysics and parameterized convection • SPC experience with simulated reflectivity from 4.5 km experimental WRF-NMM has been very positive • Higher resolution and explicit precipitation processes only • More realistic reflectivity structures/magnitudes compared to 12 km WRF • Early experience with NAM-WRF reflectivity suggests 1-hr precipitation and instantaneous precipitation rate may provide similar information as reflectivity • When displaying output with higher resolution accum. pcpn scale

  25. NAM-WRF Evaluation - Reflectivity 27 hr NAM-WRF Forecasts Valid 03z 25 May Simulated 1 km AGL Reflectivity Instantaneous Pcpn Rate

  26. NAM-WRF Evaluation - Reflectivity 27 hr NAM-WRF Forecasts Valid 03z 25 May Simulated 1 km AGL Reflectivity 1-hr Accumulated Pcpn

  27. NAM-WRF Evaluation - Precipitation 27 hr NAM-WRF Forecasts Valid 03z 25 May 3-hr Accumulated Pcpn 1-hr Accumulated Pcpn • Good news - WRF often provides more detailed pcpn structure compared to Eta • Tracks of heavier pcpn cores evident in WRF output (rarely seen in Eta)

  28. SPC NAM-WRF Evaluation • Conclusions • Low level moisture • Better in WRF (thumbs up) • Pre-convective PBL sounding structure • WRF often different but not necessarily better (thumbs sideways) • Precipitation • WRF often exhibits more detailed structures (thumbs up) • At times WRF evolution is difficult to understand (thumbs sideways) • For severe weather forecasting, WRF sometimes better; Eta sometimes better (cautious thumbs sideways) • Synoptic patterns • Generally similar, although some evidence that WRF predicts deeper 500 mb troughs compared to Eta (thumbs sideways) • SPC Recommendation – a cautious thumbs up

  29. Acknowledgements • Greg Carbin (SPC) • Jason Levit (SPC) • David Bright (SPC) • Gregg Grosshans (SPC) • John Hart (SPC) • SPC Forecasters

More Related