This directory contains the 1-degree binned monthly means and number of observations per bin used to derive the FSU3 fluxes. The data are gridded in a manner so that the grid cell is centered on the whole degree in latitude and longitude. For example, a grid with lat=1 would contain data from [0.5 to 1.5) degrees. The individual monthly netCDF files are separated into yearly directories. Within each directory you will find 3 types of monthly files: VOSYYYYMM.nc - Ship observations driftingYYYYMM.nc - Drifting buoy observations mooredYYYYMM.nc - Moored buoy observations To construct the FSU3 we separate VOS, drifting, and moored observations so that we can independently weight these data in our variational analysis. The processing procedure to create the gridded input files can be found online at: http://www.coaps.fsu.edu/RVSMDC/FSUFluxes/technical.php The present version of the moored files also has been filtered to remove any monthly mean that was derived from 4 or fewer observations. We found this to be an effective tool to eliminate mis-identified platforms (e.g., ship tracks) from the moored files. Even if a mooring was removed that was reporting 4 of fewer observations, we feel this mooring would be unrepresentative of the mean conditions over month. Missing values are set to -9999. The number of observations is set to zero when the value is missing. These files may be replaced by more recent versions without warning. An update history will be provided in this readme. Update history: The first version of the 1-degree input files was placed on line on 11 January 2008. 9 Febrary 2011: Uploaded the 1-degree input files for the years 2005-2007. These data were run using the ICOADS version 2.5 (an update since the pre-2005 years were run). ICOADS version 2.5 is stored in a new ASCII format (the international maritime meteorological archive format) which required COAPS to update our code to handle the new format and some minor modifications to the way quality flags were stored. We made every effort to match the quality flag handling used in the pre-2005 files. Otherwise, the processing of these data are identical to previous versions. 12 December 2011: Noted an error in the units for the specific humidity. Although the units listed in the netCDF files are g/kg, the actual values of specific humidity are in kg/kg. We are working to change the units in the netCDF files and will replace the files when the change is made. 23 December 2011: Units of specific humidity (spec_hum) have been corrected to kg/kg and the files have been updated on the ftp site. If you are using files downloaded prior to this date, you may wish to download the corrected files. Questions or Comments: winds@coaps.fsu.edu SRS - 12 December 2011