Summary of Blending Scheme, Version 5.0 February 27, 2008 --------------------------------------- Ocean surface wind data are derived from spatial blending of high-resolution satellite data (Seawinds instrument on the QuickSCAT satellite - QSCAT) and global weather center re-analyses (NCEP), resulting in high temporal and spatial resolution datasets (6-hourly, and 0.5 x 0.5 degree) of wind vector components u,v, and windstress curl. QSCAT provides high-wavenumber, but temporally intermittent data: each satellite revolution takes 101 min and covers a 1800 km wide swath at 25 km resolution. The NCEP fields are ubiquitous, but low-wavenumber: each global field is available every 6 hours on a T62 Gaussian grid (ca. 1.8 degrees), but the true spatial resolution is coarser than T62. The blending scheme (adapted from Chin et al., 1998, see below) creates 6-hourly global fields by retaining QSCAT wind retrievals in swath regions, and by augmenting in the unsampled regions the low-wavenumber NCEP fields with a high-wavenumber component that is based on monthly QSCAT statistics. These statistics are derived from 8 x 8 degree area averages. The augmentation preserves the observed power-law relation between power-spectral density (PSD) and wavenumber k, PSD ~ k**p, for each of the vector wind components U and V. Based on those monthly statistics, the exponent p takes values between -2 at high latitudes and -5/3 at the equator. Nearly uniform global coverage from QSCAT is achieved in 12-hr composites. Each 6-hourly analysis field from NCEP is blended with a QSCAT 12-hr composite centered on the analysis time. For highest data quality, the following data were excluded from the blending scheme: rain-flagged wind vector cells (WVC), wvc_quality_flag bit12 = 0 and bit13 = 1, and the outermost three WVC's along the edges of the satellite swath. Details of the effect of rain on QSCAT wind retrievals are described in Milliff et al., 2004. Caution ------- This blended wind product was developed for general circulation scale analyses. It should not be used when mesoscale or ultra-high resolution is required. Moreover, each 6-hourly surface wind field is derived from the latest 12-hours of QSCAT observations. This means that alternating halves of the globe retain the same QSCAT obs in each blended field output. Therefore, the blended winds are not suited for point-by-point, temporal comparisons (i.e. with buoy data), or when true 6-hourly resolution is needed. Due to seasonal ice coverage, QSCAT data poleward of 60 degree is sparse. The NCEP re-analysis are of much lower quality in polar regions as well. QSCAT winds can be contaminated near ice edges resulting in very high wind speeds that are not always removed by quality control. Therefore, blended winds poleward of 60 degree latitude are of reduced quality, and gradient wind products, such as wind stress curl and divergence, cannot be trusted. QSCAT data ---------- Satellite data are derived from the NASA Quick Scatterometer (QSCAT) mission and are based on the DIRTH algorithm ("Direction Interval Retrieval with Threshold Nudging"). The current version of QSCAT data is based on the latest release of scatterometer data. A complete reprocessing of QSCAT was finished in March 2007 and is based on a recalibrated model function called QSCAT-1/F13, resulting in slightly higher wind speeds above 16m/s. In addition, a new rain-flagging technique is employed, called "IMUDH" (Impact-based Multidimensional Histogram). Compared with the previous algorithm, this technique does not flag a specific rain rate, but instead, a likelihood that wind speeds or directions are perturbed by rain. As a result, fewer data are rain-flagged. NCEP data --------- The re-analyses winds are from the NCEP/NCAR Reanalysis project, and the data are available from: http://www.cdc.noaa.gov/cdc/reanalysis/. The reference is: Kalnay et al., "The NCEP/NCAR 40-year Reanalysis Project", Bull. Amer. Meteor. Soc., 77, 437-470, 1996. Earlier Versions ---------------- Version 1.0: All rainflagged QSCAT wind vector data were excluded from blending. Version 2.0: Only rain-flagged data with wind speeds less than 15 m/s are excluded. Recent work demonstrates a wind-stress curl bias due to rain-flagged WVC in high-wind speed frontal and low-pressure systems (Milliff et al., 2004). Buoy comparisons for wind speeds greater than or equal to 15 m/s suggest that the rain contamination is overwhelmed by radar backscatter effects due to wind at those speeds. Version 3.0: Improves on the computation of the monthly statistics in 8 x 8 degree bins, by using a more even geographic distribution of samples. Version 4.0: Corrects some artifically high winds in previous versions that resulted from a) temporal mismatch of up to 6 hrs in the blending of QSCAT swath data with background NCEP winds; and b) splining of unrealistic QSCAT winds exceeding 15 m/s in rain with NCEP winds. Version 5.0: Incorporates the latest release of QSCAT with new rain-flagging. In the current blending scheme, all rain-flagged data are excluded, regardless of wind speed. Older version winds are available upon request (please contact the data specialist listed on the SET HOME page) Geographic Grid --------------- The data domain is a 0.5x0.5 degree grid, 720 x 353, starting at (0.5E, 88.0S) and ending at (360.0E, 88.0N). There are data values at all grid points, including land and ice points. Users who wish to examine ocean points need to use appropriate land/ice masks. The blended wind product was designed to provide valid winds over the ocean; land values are simply "filled-in" and are not to be used as realistic values. Land Mask --------- A land mask is available, and can be read with software provided on the "SET SOFTWARE" web page. The data file, "mask_blended_winds.v02.dat", is an ASCII text file, where land is marked with "1", and ocean with "0". It contains the dimension of the mask (720 x 353), and the longitude and latitude grid values. This land mask, however, should be used only after careful examination and possible editing depending on the desired result. The land mask is based on a 5min topography data set that was obtained from NCAR. It was created to eliminate land and coastal values in a very conservative manner. For each 0.5 x 0.5 degree blended wind grid cell, there are 6x6=36 (5min) topo data points. Whenever there is at least one 5min land point within a 0.5 x 0.5 grid cell, the mask was set to the land value. Furthermore, any adjacent "coastal" points were also set to the land value. All inland seas, like the Black and Caspian Seas, and some marginal seas, like the Red Sea and the Persian Gulf, are set to the land value. CAUTION: Some very small islands may not be masked in this file. These may have fairly tall mountains that deflect the surface winds. Users who are not interested in these small-scale features, need to edit the mask to suit their needs. Time convention --------------- Time is 6-hourly. The "day" variable in the data files is in "days of year": for example January 1, 6hr = day 0.25. There is no February 29 in leap years. This data was produced for ocean models that required a constant number of days per year. In all years the day number goes from 0.25 to 365.00. References ---------- The details of the blending methology are the subject of "Basin-Scale High-Wavenumber Sea Surface Wind Fields from Multiresolution Analysis of Scatterometer Data", Chin, T.M., R.F.Milliff, and W.G.Large, 1998, J. Atmos. Ocean. Technology, 15, 741-763. A blended wind product derived with NSCAT scatterometer data (covering September 1996 to June 1997) is described in the appendix of "Ocean General Circulation Model Sensitivity to Forcing from Scatterometer Winds", Milliff, R.F., W.G.Large, J.Morzel, G.Danabasoglu, and T.M.Chin, 1999, J. Geophys. Res., C5, 11337-11358. The rain contamination of scatterometer wind retrievals are described in "Wind Stress Curl and Wind Stress Divergence Biases from Rain Effects on QSCAT Surface Wind Retrievals", Milliff, R.F., J.Morzel, D.B.Chelton, and M.H.Freilich, 2004, J. Atmos. Ocean. Tech., 21, 1216-1231.