Extended Abstract (2.2M)10.5
Operational evaluation of QuikSCAT ocean surface vector winds in tropical cyclones at the Tropical Prediction Center/National Hurricane Center
Michael J. Brennan, UCAR/TPC/NHC, Miami, FL; and R. D. Knabb
Remotely sensed ocean surface vector winds from the QuikSCAT scatterometer aboard the NASA SeaWinds satellite are an important source of data at the Tropical Prediction Center (TPC)/National Hurricane Center (NHC). These data fill a large void in surface wind information over the open ocean portion of the NHC/TPC area of responsibility in the Eastern Pacific and Atlantic basins. QuikSCAT has been used at TPC/NHC to supplement other observations for estimation of tropical cyclone (TC) intensity and the radius of 34-kt and 50-kt winds. Both of these parameters are critical to the warning and forecast mission at TPC.
In the TC environment, the quality of QuikSCAT winds is often degraded by rain contamination. However, the interpretation of the impact of rain on QuikSCAT winds is not straightforward, since no direct measurement of rain rate is co-located with the instrument, and the impact of rain varies with the rain rate and the actual near-surface wind speed. Examination of QuikSCAT wind data within the range of coastal radar and other rain rate estimates such as those from the Tropical Rainfall Measurement Mission (TRMM) have shown a strong relationship between high wind estimates from the scatterometer and rain rate.
To quantify the accuracy of QuikSCAT winds in estimating TC intensity, maximum wind speeds from 25-km and 12.5-km near-real time QuikSCAT retrievals over classified TCs during the 2005 and 2006 Atlantic hurricane seasons are compared to the NHC best track intensity and Dvorak satellite intensity estimates. Results from 2005 indicate a positive bias in QuikSCAT maximum winds for tropical depressions and weak tropical storms due to rain enhancing backscatter to the instrument. A large negative bias of QuikSCAT intensity estimates was seen for hurricanes with maximum winds above 85 kt due to rain attenuation, resolution limitations, and signal saturation at high winds speed. QuikSCAT intensity estimates showed more skill in moderate to strong tropical storms and marginal hurricanes; although some of this reduced bias results from the cancellation of larger errors of opposite sign in this range of wind speeds. Estimates of outer wind radii have proven very useful, especially in major hurricanes where the maximum radius of 34 kt winds often extends outside the region of heavy rainfall near the TC center.
These bulk statistical results indicate that QuikSCAT winds in the TC environment can provide valuable information, but must be carefully evaluated on a case-by-case basis. Results from 2006 will be evaluated and compared to those from 2005. Also, case examples comparing QuikSCAT winds to measurements from aircraft reconnaissance, the NOAA stepped-frequency microwave radiometer (SFMR), and ship and buoy observations will be presented. These examples will highlight the difficulty of interpreting QuikSCAT winds in an operational setting. Additionally, an evaluation of a new near-real time QuikSCAT retrieval algorithm developed by NASA and NOAA/NESDIS may be possible prior to the end of the 2006 hurricane season. Finally, suggestions will be made for improvements to future ocean vector wind platforms necessary to support the operational mission at TPC.
Session 10, Land and Ocean Observations
Thursday, 18 January 2007, 11:00 AM-2:30 PM, 212B
Previous paper Next paper
Recorded presentation