TU1.T10.2.pptx

363 views
319 views

Published on

Published in: Technology, Business
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
363
On SlideShare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
4
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide
  • Put markers every 90 degrees for the GMF plots to differentiate the two GMFs
  • Put markers every 90 degrees for the GMF plots to differentiate the two GMFs
  • X-Winds & L2B color scatter plots
  • X-Winds & L2B color scatter plots
  • TU1.T10.2.pptx

    1. 1. A Novel Ocean Vector Winds RetrievalTechnique for Tropical CyclonesPeth Laupattarakasem1, Suleiman Alsweiss1, W. Linwood Jones1, and Christopher C. Hennon21Central Florida Remote Sensing LaboratoryUniversity of Central FloridaOrlando, Florida 328162University of North Carolina AshevilleAsheville, NC 28804IGARSS 2011Vancouver, CanadaJuly 26, 2011<br />
    2. 2. Outline<br />SeaWinds OVW measurements in hurricanes<br />Radar backscatter geophysical model function<br />X-Winds OVW retrieval algorithm<br />wind direction retrieval<br />wind speed retrieval<br />X-Wind Comparisons with H*Wind & QuikSCAT standard L2B OVW product<br />
    3. 3. SeaWinds Hurricane Measurements<br />Historically Ku-band scatterometers have under estimated hurricane wind speeds<br />Issues<br />Inadequate spatial resolution (25 km IFOV)<br />Rain contamination (backscatter & attenuation)<br />Geophysical OVW retrieval algorithm<br />Developed for global winds<br />Not optimized for tropical cyclones<br />
    4. 4. Extreme Winds Sigma-0 Geophys Model Function (XW-GMF)<br />Special QuikSCATGMF developed for hurricanes<br />Training dataset of 35 QScat hurricane overpasses<br />3-D GMF: so = f(ws, relative wdir, atmostransmis)<br />WS: one-minute sustained 10m wind speeds from NOAA HRD H*Wind surface wind analysis<br />Relative wind direction c:the difference between wind direction and the radar azimuth look<br />Atmostransmissivity: inferred from simultaneous observations of QRad H-pol brightness temperatures<br />
    5. 5. Hurricane Geophysical Model Function (GMF) <br />30 m/s<br />H-pol so, dB <br />
    6. 6. GMF C0- CoeffDependence on Wind Speed<br /> C0 , dB <br />10 20 40 60<br /> Wind Speed, m/s <br />
    7. 7. GMF Anisotropy with Wind Speed<br /> C1 , linear units <br /> C2 , linear units <br />10 20 40 60<br /> Wind Speed, m/s <br />10 20 40 60<br /> Wind Speed, m/s <br />
    8. 8. XW-GMF Atmospheric Transmissivity<br />Rain attenuation is corrected implicitly through use of the QRad H-pol brightness temperature Tbh<br />GMF = f(ws, rel_wdir, Tbh)<br />Tbh<br />140 k<br />160 k<br />190 k<br />
    9. 9. X-Winds OVW retrieval Algorithm<br />Attributes<br />Assumes cyclonic wind direction rotation about TC center<br />Assumes rain effects are primarily absorptive<br />Rain volume backscatter is neglected<br />Empirical 3-D XW-GMF accounts for backscatter saturation with wind speed & rain absorption<br />Uses sequential scalar wind direction and wind speed estimation<br />Not traditional ocean vector wind “maximum likelihood estimation” retrieval <br />
    10. 10. Scalar Wind Direction Estimation<br />Relies on anisotropy of measured difference between forward and aft looking backscatter measurements<br />Dsomeas = (sofore – soaft) @ top-of-the-atmos<br />Typical hurricane image<br />Latitude Index<br />Dso - linear units<br />Longitude Index<br />
    11. 11. Wind Direction Modeling of Dso<br />so anisotropy model for single radar azimuth look<br />Taking the difference between fore & aft radar looks yields<br />Wind direction solutions (aliases) are roots of<br />Yields ambiguous wind direction solutions <br /> typically ~ 6 – 8<br />
    12. 12. Wind Direction Alias Removal<br />Keep ambiguities within window ±30 from CCW spiral<br />Multi-pass median filter and populate missing pixels through interpolation<br />Use smoothed TC wind field for wind speed estimation<br />Wind Directions Ambiguities Retrieved Wind Direction Field<br />
    13. 13. X-Winds Scalar Wind Speed Estimation<br />Use smooth TC wind field as “estimated true” wind direction and calculate relative wind direction <br /> c = (radar az – “est-wind dir”)<br /><ul><li>Find 1-D wind speed solution for each “so flavor”
    14. 14. (V- & H-pol, fore & aft direction) that satisfies this relationship</li></li></ul><li>X-Winds Wind Speeds Retrievals <br />|(someas – GMF)|<br />4 σ0 flavors <br />HF HA VF VA<br />Wind Speed (m/s)<br />
    15. 15. Results and Performance Evaluation<br />X-Winds OVW retrievals<br />- L2B-12.5km OVW product- H*Wind surface wind analysis<br />
    16. 16. Hurricane Fabian Rev#21898<br />X-Winds <br />L2B-WS 12.5km<br />H*Wind Analysis<br />
    17. 17. Hurricane Ivan Rev#27217<br />X-Winds <br />L2B-WS 12.5km<br />H*Wind <br />
    18. 18. Wind Speeds Comparison (10 revs)<br />X-Winds QSCAT L2B-12.5km<br />TbH-pol<br />TbH-pol<br />
    19. 19. Wind Speeds Comparison (10 revs)<br />X-Winds QSCAT L2B-12.5km<br />TbH-pol<br />TbH-pol<br />
    20. 20. Wind Direction Comparison (10 revs)<br />X-Winds QSCAT L2B-12.5km<br />TbH-pol<br />TbH-pol<br />
    21. 21. Wind Direction Comparison (10 revs)<br />X-Winds QSCAT L2B-12.5km<br />TbH-pol<br />TbH-pol<br />
    22. 22. Conclusion<br />A new OVW retrieval algorithm has been developed for SeaWinds<br />Specifically tailored for tropical and extra-tropical cyclones<br />Performance evaluated using comparisons with H*Wind surface wind analysis and L2B-12.5km<br />Better agreement with H*Wind<br />A new (11-year) SeaWindsTC OVW data set will be produced starting next year 2012<br />

    ×