low-level wind analysis and prediction during b08fdp 2006
DESCRIPTION
5th ICMCS - Oct. 31, 2006. Low-level Wind Analysis and Prediction During B08FDP 2006. Juanzhen Sun and Mingxuan Chen. Other contributors: Jim Wilson Rita Roberts Sue Dettling Yingchun Wang Xiaoguang Tan Feng Gao. Outline. Description of the analysis system VDRAS - PowerPoint PPT PresentationTRANSCRIPT
Low-level Wind Analysis and Prediction During B08FDP 2006
Juanzhen Sun and Mingxuan Chen
Other contributors:Jim WilsonRita RobertsSue DettlingYingchun WangXiaoguang TanFeng Gao
5th ICMCS - Oct. 31, 2006
Outline
• Description of the analysis system VDRAS• Preliminary results from B08FDP 2006• Summary and future work
5th ICMCS - Oct. 31, 2006
What is VDRAS?
5th ICMCS - Oct. 31, 2006
• Continuously cycled 4D-Var system of a cloud-scale model
• Assimilate both radial velocity and reflectivity
• Can be updated every 12 min.
• Produce analyses of wind, temperature perturbation,
relative humidity, and some microphysics
• Has been implemented in real time at more than ten
locations
Variational Doppler Radar Analysis System
Flow chart of VDRAS analysis procedures
4DVar constrained by a cloud-scale model
radial velocityreflectivity
u,v,w,T, qr,qc ,qv
Mesoscale analysisMM5 output
Surface obs. VAD analysis
0 min 6 min 12 min
Preprocessing andQuality control
Integrated display with
ANC
Background
5th ICMCS - Oct. 31, 2006
4DVar radar data assimilation
VDRAS Verification from previous studies
• ACARS (Sun and Crook 2000)
• Dual-Doppler (Crook and Sun 2004)
• Research aircraft (Sun and Crook 1998)
5th ICMCS - Oct. 31, 2006
Cpol
Kurnell
rms(udual – uvdras) = 1.4 m/s
rms(vdual – vvdras) = 0.8 m/s
Dual-Doppler verification
VDRAS implementation in Beijing
• Domain size: 100X100X14 grid points with a 3kmx3kmx375m resolution assimilating BJRS and TJRS data
• Run on a 2-processor Dell workstation
• Take ~ 20 min for each analysis (with a 12 min assimilation window) plus a 1-hour forecast
5th ICMCS - Oct. 31, 2006
Special Issues for VDRAS in Beijing
5th ICMCS - Oct. 31, 2006
• Complex terrain
• Radar data quality - AP problem - Radars only operate in storm
mode • Dense surface network in
Beijing but sparse outside
• MM5 analysis only twice a day
Squall line of July 10, 2006
0408 UTC 0550 UTC
Convective system enhanced by southeast flow
VDRAS analysis
Wind and Convergence at z = 0.938 km Wind and relative humidity at z = 0.563 km
0408 UTC 0408 UTC
Severe storm of Aug. 1, 2006
Storm initiated on the mountain and enhanced by low-level boundary interaction
0904 UTC
Observed reflectivity Analyzed convergenceZ = 0.9325 km
0925 UTC0947 UTC1007 UTC1029 UTC
1-hour forecast winds overlaid on observed reflectivityupdated every 20 min.
Summary and Future Work
5th ICMCS - Oct. 31, 2006
• VDRAS was able to run robustly at BMB, assimilating radial velocity and reflectivity from BJRS and TJRS• VDRAS produced analyses on a 300kmx300km domain every 20
min.• Preliminary case studies indicated that VDRAS produced
reasonable analysis even with the complex terrain• Further studies are needed to verify VDRAS analysis using independent data• Surface wind (2 meter above ground) forecast product will be
generated for water sports venues