1 / 32

Challenges in TC Modeling, Predition, and Data Assimilation

Challenges in TC Modeling, Predition, and Data Assimilation. Shuyi S. Chen Rosenstiel School of Marine and Atmospheric Science University of Miami NEXRAD In Space (NIS) Workshop, Miami, 10-11 April 2007. Track forecast errors cut in half in 15 years. From Richard Knabb.

ailis
Télécharger la présentation

Challenges in TC Modeling, Predition, and Data Assimilation

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. Challenges in TC Modeling, Predition, and Data Assimilation Shuyi S. Chen Rosenstiel School of Marine and Atmospheric Science University of Miami NEXRAD In Space (NIS) Workshop, Miami, 10-11 April 2007

  2. Track forecast errors cut in half in 15 years From Richard Knabb

  3. No progress with intensity NHC Official Intensity Error Trend From Richard Knabb

  4. TC Formation/Genesis • Very little skill in current operation models • Not fully tested using high-resolution research model • TC Structure and Intensity Change • Inner core (eye and eyewall) dynamics and rainbands • Environmental conditions, including vertical wind shear, moisture distribution, and sea surface temperature (upper ocean heat content), surface properties (e.g., waves, sea spray, etc.) • TC Landfall • Land surface, TC – trough/front interactions, etc.

  5. What is a “next-generation” hurricane model? • High-resolution to resolve inner core and rainband structures (~1 km grid spacing or better?) • Fully coupled atmosphere-wave-ocean-land modeling system • Data assimilation system • Both deterministic and probabilistic forecasting capabilities

  6. High-Resolution Multi-nested Vortex-Following Numerical Models: 15 km • UM/RSMAS Coupled Atmos-Wave-Ocean Model Mini ensemble MM5 and WRF forecasts using GFS, NOGAPS, CMC, and GFDL forecast fields as initial and lateral boundary conditions 5 km 1.6 km

  7. MM5 Hurricane Floyd (1999)

  8. Obs NHC Fcst MM5 (1.6km) WRF (1.6km)

  9. MM5&WRF Global models NHC Fcst Obs

  10. Houze, Chen, Smull. Lee, and Bell (2007) Figure 1: High Resolution Model Output for Rita Forecast rain rate in Hurricane Rita during RAINEX

  11. Figure 2: Radar analysis

  12. Confirmation of concentric eyewalls in Hurricane Rita (2005) N43 flight-level wind in Rita Inner outer outer Inner Observed MM5 22 September 2005 Radar Reflectivity

  13. Effect of vertical wind shear on Rita Model 5-day forecast

  14. Sensitivity to Model Resolution

  15. Rita Rita 1.6 km 5 km

  16. Hurricane Rita Large initial vortex

  17. Hurricane Rita 5 km 1.6 km vorticity wind 1.6 km 5 km IS = [(f + 2V/r)(f + V/r + V / r)] 1/2

  18. Hurricane Floyd 5 km 1.67 km

  19. 5 km (25 km2) 1.67 km (2.8 km2)

  20. Sensitivity to Physical Parameterizations

  21. Microphysics Param in WRF (Tao et al 2007) PBL Param in MM5 (Braun&Tao 2000) Air-Sea Coupling (Chen et al 2007) Sea Spray Param (Desflots et al 2007)

  22. Coupled Atmosphere-Wave-Ocean Modeling System for Hurricane Predictions ATMOS. MODEL (MM5/COAMPS/WRF) OCEAN MODEL (HYCOM/3DPWP/ROMS) Surface fluxes SST Param. of wave dissipation LES Sea Spray SSH & current velocity Wind-induced stress Wave-Induced stress Roughness length WAVE MODEL (WAVEWATCH III/WAM) Param. of spectral tail and drag coefficient Source function? Drop size distribution? Effects on turbulence? How do these affect exchange coefficients? What is the ratio of Ck and CD?

  23. CBLAST

  24. WRF/EnKF for Katrina with Radar DataFuqing Zhang, Texas A&M University • EnKF: as MZ07a but with 30-member ensemble, no physics uncertainty • WRF Domains 1-4: 40.5, 13.5, 4.5 &1.5km (2 way); 35 vertical levels • WSM 6-class microphysics scheme w/ graupel • YSU boundary-layer scheme • Grell-Devenyi ensemble cumulus scheme • IC & BC: FNL analysis • Data assimilated: • TC position (err=20km) and intensity (err=1hpa) D1,D2 every 3 h from 0Z/25 to 0Z/26 (Chen and Snyder 2006) • Doppler velocity (err=3m/s) from KMAX and KBYX (D1,D2) at 00Z/26 and/or 03 & 06Z • Doppler velocity from airborne radar at 18Z/27 (err=3m/s) for all domains down to 1.5-km D4 262x253x4.5km 253x 253 x 13.5km 160x 120 x 40.5km

  25. EnKF Performance: Track forecast with Vr assimilated NEXRAD at KAMX and KBYZ assimilate at 00Z 26 Aug in 14.5-km D2; free forecast with 4.5-km D3 afterwards Airborne Radar assimilated at 18Z 27 Aug at all domains down to 1.5-km domain; free forecast afterwards

  26. EnKF Performance: Maximum Surface Wind Simulations NEXRAD at KAMX and KBYZ assimilate at 00Z 26 Aug in 14.5-km D2; free forecast with 4.5-km D3 afterwards Airborne Radar assimilated at 18Z 27 Aug at all domains down to 1.5-km domain; free forecast afterwards

  27. Challenges • TC structure and intensity are very sensitive to model grid resolution, surface fluxes, boundary layer and microphysical parameterizations • Lack of good model initial conditions • Lack of high-resolution observations to evaluate/verify coupled model forecasts • Need robust and efficient data assimilation system • What is the Impact of assimilating small-scale features on forecasts longer than 2-3 days?

  28. Questions • What are the forecasting products needed from the next-generation models (e.g., not just track and Vmax, but also storm impact including rain, wind, wave, storm surge, etc.)? • How do we develop a evaluation/validation system for high-resolution products (e.g., wind and rainfall distributions, storm surges)? • How will these be used operationally, perhaps a combination of deterministic and probabilistic measure that takes into account of track errors?

  29. Model Forecast of Hurricane Katrina (2005) (water depth) RainRate Surface Wind Speed (m/s) Significant Wave Height (m)

More Related