The Quicklook Objective FSU Tropical Pacific Winds - a description

***NOTE: As of October 2020, the source GTS data for the FSU winds has
been changed from NCEP to the Observing System Monitoring Center (OSMC). See 
http://osmc.noaa.gov/erddap/tabledap/OSMC_30day.html for dataset
details. FSU verified that the two GTS sources are very similar; 
however the OSMC does include more wind records (especially for western
Pacific moorings).***

Version 1.1
7 October 2002

The Center for Ocean-Atmospheric Prediction studies at FSU is pleased 
to release verion 1.1 of our objectively analyzed quicklook fields of 
tropical Pacific pseudo-stress. The data are available January through
September of 2002 on this ftp site. Susequent objective quicklook fields
will be created monthly, typically in the first week of the month for the
the previous month (e.g., Oct. 2002 will be available in early Nov. 2002).
The netcdf files are self describing, while the ascii files are outlined below.

NOTE FOR USERS OF THE OLD SUBJECTIVE FSU WINDS: please note that the objective
product is on an 85 by 30 grid. We added values centered on 122E to bridge the
gap between our tropical Pacific and tropical Indian ocean products.

IMPROVEMENTS SINCE BETA VERSION 0.5:
1. Improvements have been made to the land mask to address known problem 2 
below.

2. Buoy data have been added to the background field. Earlier version only 
used ship data.

3. Weighting function for the background has been modified.

KNOWN PROBLEMS: 
1. Sampling-related issue cause the uncertainty (to be provided in a future 
release) to be non-uniform across the field. Note that this is true of any 
other publicly available product of which we are aware. Greater caution 
should be applied when drawing conclusions about data-void areas, such as 
the south eastern Pacific Ocean.

2. The improved land mask still results in a small number of relatively poor
vectors in regions with many islands (e.g., Indonesia). This 
is a consequence of not averaging across land (which usually results in 
more accurate winds). The problem is related to poor sampling and will 
be addressed in a later update.

3) The lower data density in the real-time GTS data from NCEP still results in
problems south of Panama and north of New Guinea. Tests are underway to
address this problem and fixes will be made by early 2003.

The netCDF (.nc) and ASCII (.asc) files contain Pacific Ocean pseudostress from
in-situ ship and buoy reports on an 85x30 grid, 1 grid/month. Monthly netCDF
file are: FSU_obj_psv_qpacYYYYMMv11.nc and ASCII files are: 
FSU_obj_psv_qpacYYYYMMv11.asc (where YYYY is the 4-digit year and MM is the 
2-digit month). The analyses are a quicklook product based on real-time GTS 
surface marine observations from NCEP. The fields should be treated as 
preliminary in nature and will be replaced annually with a research quality
product based on marine data from the National Climatic Data Center 
(NCDC). PMEL TAO buoys are included in these analyses. Note: no satellite data
are included in this monthly product.

The pseudostress (tau x, tau y below) are in (meters/sec)**2 with east = 
positive and north = positive. Further information on conversion to other units
is available via the web (http://www.coaps.fsu.edu/~bourassa/scat_html/
forcing_tut/forcing_tutorial.html#pseudostress).

Any tau x and tau y with a value of 999.0 is fully over land.  The data are
on a 2 by 2 degree grid, with 85 points e-w and 30 points n-s. The designation
of cells as sea or land is based on the ElTopo5 topography. Any grid cells with 
eltopo cell elevations (from a twelfth of a degree grid) less than mean sea 
level are treated as water. The few inland points that meet this condition are 
treated as land. Observations from land cell do not contribute to our gridded 
product.

NOTE: THE NEW OBJECTIVE PRODUCT IS AVAILABLE FROM 122E TO 70W LONGITUDE,
WHEREAS THE OLD SUBJECTIVE FSU WIND GRID WAS FROM 124E TO 70W LONGITUDE.

       tau x( 1, 1) and tau y( 1, 1) are centered at 122e and 29s 
       tau x( 2, 1)    "     "       are centered at 124e and 29s 
         .
         .
       tau x(84, 1)    "     "       are centered at  72w and 29s 
       tau x(85, 1)    "     "       are centered at  70w and 29s 


       tau x( 1, 2)    "     "       are centered at 122e and 27s 
       tau x( 1, 3)    "     "       are centered at 122e and 25s 
         .
         .
       tau x( 1,29)    "     "       are centered at 122e and 27n 
       tau x( 1,30)    "     "       are centered at 122e and 29n 

The following FORTRAN code is an example of reading the data. Note: the
data in the files are scaled up by a factor of 10, and that using this
format will scale them to the proper numbers.


      program readwd
      dimension taux(85,30),tauy(85,30) 

c     read 1 month wind stress grid

      read (1,10) month,iyear,taux,tauy 
   10 format (2i5,14f5.1,/(16f5.1)) 
       .
       .
       .
       .
       .
      end 

We acknowledge the support of NOAA Office of Global Programs and the National
Science Foundation for funding this project.
      
REFERENCE: 

Bourassa, M. A., S. R. Smith, and J. J. O'Brien, 2001: A new FSU winds and
flux climatology. 11th Conference on Interations of the Sea and Atmosphere,
San Diego, CA, Amer. Meteor. Soc., 912.

CONTACT:

Shawn R. Smith
COAPS
The Florida State University
Tallahassee, FL  32306-2840

Voice: (850) 644-6918
Fax:   (850) 644-4841
Internet:winds@coaps.fsu.edu