1 Proceedings OceanSAR 2006 – Third Workshop on Coastal and Marine Applications of SAR, St. John’s, NL, Canada, October 2006 SAR-derived Winds from Hurricanes: Assimilative Blending with Weather Forecast Winds W. Perrie a , W. Zhang b , M. Bourassa c , H. Shen a,d and P.W. Vachon e a BIO, Fisheries and Oceans Canada, Dartmouth, Canada, email: perriew@dfo-mpo.gc.ca b Environment Canada, Downsview, Ontario, Canada c Center for Ocean-Atmospheric Prediction Studies, The Florida State Univ., Tallahassee, FL d Institute of Oceanology, Chinese Academy of Sciences, Qingdao, Shandong, China, e Defence R&D Canada - Ottawa ABSTRACT We use a variational data assimilation method to combine remotely sensed SAR (synthetic aperture radar) wind data from Hurricane Isabel (2003) with other sources of winds, including scatterometer, and a blended surface wind analysis combining QSCAT and the NCEP reanalysis. Resulting winds are validated with in situ buoy observations. We show that the newly-constructed winds are consistent with independent observations. Preliminary results suggest they are an improvement over QSCAT/NCEP blended winds. Keywords: SAR-derived winds, Hurricane Isabel, variational data assimilation. 1 INTRODUCTION In September 2003, Hurricane Isabel developed to category 5 intensity north of the Caribbean, and moving toward the northwest, weakened to category 2 as it made landfall in North Carolina. To generate high-quality marine wind estimates for this hurricane, we start with RADARSAT-1 SAR-derived wind data, and generate new marine winds by using a variational data assimilation method to combine the SAR data with background fields consisting of QuikSCAT (QSCAT), or alternatively, QSCAT/NCEP (National Centers for Environmental Prediction) analysis winds. Resultant winds are shown to compare favorably with in situ buoy winds relative to what is achieved with the QSCAT/NCEP winds. The variational method uses constraints to create an optimum combination of winds, in the sense of minimization of a cost function measuring the misfit between observations and background input field data, and constraining non-geophysical features in the spatial derivatives. QSCAT/NCEP winds are obtained by blending QuikSCAT scatterometer data with NCEP reanalysis winds [Chin et al., 1998; http://dss.ucar.edu/datasets/ds744.4/ ]. Section 2 briefly describes Hurricane Isabel and the methodology. Section 3 gives marine wind results and Section 4 gives conclusions. 2 HURRICANE ISABEL Hurricane Isabel was first categorized as a Tropical Storm in the central Atlantic Ocean on 6 September 2003. It intensified to a category 5 hurricane and moved westward, north of the Caribbean and Bahamas. On 14 September, Isabel’s movement slowed, after turning towards the northwest. Cooler waters over which it passed prevented further intensification. It continued to weaken as it headed toward the North Carolina coast, making landfall as a category 2 hurricane at Ocracoke Island between Cape Hatteras and Cape Lookout, with maximum sustained winds near 85 kts around 17 UTC on 18 September. Our variational methodology [Pegion et al., 2000; O’Brien et al., 2003; Bourassa et al., 2002; Morey et al., 2005] is an objective technique based on minimization of a cost function, which measures misfit to observations, and smoothing with respect to several background fields. The influence of a background field is controlled by weights on the smoothing constraints. These are objectively derived by a cross validation