The Operational Use of QuikSCAT Ocean Surface Vector Winds at the National Hurricane Center

The utility and shortcomings of near-real-time ocean surface vector wind retrievals from the NASA Quick Scatterometer (QuikSCAT) in operational forecast and analysis activities at the National Hurricane Center (NHC) are described. The use of QuikSCAT data in tropical cyclone (TC) analysis and forecasting for center location/identification, intensity (maximum sustained wind) estimation, and analysis of outer wind radii is presented, along with shortcomings of the data due to the effects of rain contamination and wind direction uncertainties. Automated QuikSCAT solutions in TCs often fail to show a closed circulation, and those that do are often biased to the southwest of the NHC best-track position. QuikSCAT winds show the greatest skill in TC intensity estimation in moderate to strong tropical storms. In tropical depressions, a positive bias in QuikSCAT winds is seen due to enhanced backscatter by rain, while in major hurricanes rain attenuation, resolution, and signal saturation result in a large negative bias in QuikSCAT intensity estimates. QuikSCAT wind data help overcome the large surface data void in the analysis and forecast area of NHC’s Tropical Analysis and Forecast Branch (TAFB). These data have resulted in improved analyses of surface features, better definition of high wind areas, and improved forecasts of high-wind events. The development of a climatology of gap wind events in the Gulf of Tehuantepec has been possible due to QuikSCAT wind data in a largely data-void region. The shortcomings of ocean surface vector winds from QuikSCAT in the operational environment at NHC are described, along with requirements for future ocean surface vector wind missions. These include improvements in the timeliness and quality of the data, increasing the wind speed range over which the data are reliable, and decreasing the impact of rain to allow for accurate retrievals in all-weather conditions.

The Impact of Variational Assimilation of SSM/I and QuikSCAT Satellite Observations on the Numerical Simulation of Indian Ocean Tropical Cyclones

In this study, the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) with three-dimensional variational data assimilation (3DVAR) is utilized to investigate the influence of Special Sensor Microwave Imager (SSM/I) and Quick Scatterometer (QuikSCAT) observations on the prediction of an Indian Ocean tropical cyclone. The 3DVAR sensitivity runs were conducted separately with QuikSCAT wind vectors, SSM/I wind speeds, and total precipitable water (TPW) to investigate their individual impact on cyclone intensity and track. The Orissa supercyclone over the Bay of Bengal during October 1999 was used for simulation and assimilation experiments. Assimilation of the QuikSCAT wind vector improves the initial position of the cyclone’s center with a position error of 33 km, which was 163 km in the background analysis. Incorporation of QuikSCAT winds was found to increase the air–sea heat fluxes over the cyclonic region, which resulted in the improved simulated intensity when compared with the simulation made without QuikSCAT winds in the initial conditions. The cyclone track improved significantly with assimilation of QuikSCAT wind vectors. The track improvement resulted from relocation of the initial cyclonic vortex after assimilation of QuikSCAT wind vectors. Like QuikSCAT, assimilation of SSM/I wind speeds strengthened the cyclonic circulation in the initial conditions. This increase in the low-level wind speeds enhanced the air–sea exchange processes and improved the simulated intensity of the cyclone. The lack of information about the wind direction from SSM/I prevented it from making much of an impact on track prediction. As compared to the first guess, assimilation of the SSM/I TPW shows a moistening of the lower troposphere over most of the Bay of Bengal except over the central region of the cyclone, where the assimilation of SSM/I TPW reduces the lower-tropospheric moisture. This decrease of moisture in the TPW assimilation experiment resulted in a weak cyclone intensity.

The Impact of Assimilating SSM/I and QuikSCAT Satellite Winds on Hurricane Isidore Simulations

Three observational datasets of Hurricane Isidore (in 2002) were analyzed and compared: the Special Sensor Microwave Imager (SSM/I), the Quick Scatterometer (QuikSCAT) winds, and dropsonde winds. SSM/I and QuikSCAT winds were on average about 1.9 and 0.3 m s−1 stronger, respectively, than dropsonde winds. With more than 20 000 points of data, SSM/I wind speed was about 2.2 m s−1 stronger than QuikSCAT. Comparison of the wind direction observed by QuikSCAT with those from the dropsondes showed that the quality of QuikSCAT data is good. The effect of assimilating SSM/I wind speeds and/or QuikSCAT wind vectors for the analysis of Hurricane Isidore was assessed using the fifth-generation Pennsylvania State University–National Center for Atmospheric Research (PSU–NCAR) Mesoscale Model (MM5) and its three-dimensional variational data assimilation system. For the Hurricane Isidore case study, it was found that the assimilation of either satellite winds strengthened the cyclonic circulation in the analysis. However, the increment of the QuikSCAT wind analysis is more complicated than that from the SSM/I analysis due to the correction of the storm location, a positive result from the assimilation of wind vectors. The increase in low-level wind speeds enhanced the air–sea interaction processes and improved the simulated intensity for Isidore. In addition, the storm structure was better simulated. Assimilation of QuikSCAT wind vectors clearly improved simulation of the storm track, in particular during the later period of the simulation, but lack of information about the wind direction from SSM/I data prevented it from having much of an effect. Assessing the assimilation of QuikSCAT wind speed versus wind vector data confirmed this hypothesis. The track improvement partially resulted from the relocation of the storm’s initial position after assimilation of the wind vectors. For this case study, it was found that the assimilation of SSM/I or QuikSCAT data had the greatest impact on the Hurricane Isidore simulation during the first 2 days.