1-degree SST dataset compiled from ICOADS from 1850 to 2002

Shoshiro Minobe, Graduate School of Science, Hokkaido University, Sapporo, 060-0810, Japan

One-degree resolution SST dataset is produced using the data colleced ICOADS, without spatial interpolation nor smoothing. The SST data were corrected for biases arising from bucket and enjine measurements, with the corrected value of Parker and Folland (1995) were applied. An realistic features found in the gridded SST anomalies in ICOADS (at least in the version in April 2004) or COADS were detected and removed. Detailed gridding methodology and applications of this dataset for analysis of oceanic front variability in the Northern Hemisphere can be found in Minobe and Maeda (2005). This dataset can provide fine structures of SST variability on decadal timescales, because spatial interpolation and smooting were not applied. On the other hand, the dataset is noisy, but as shown in Minobe and Maeda (2005), one can still find a way to extract useful information from this dataset.

You can use this 1-degree resolution SST dataset for education and scientific use (not for comercial use), with proper citation and aknowledgements. Also, we would like to encourage to e-mail to me (Shoshiro Minobe ), if you download this dataset, so that I can inform you future corrections or updates.

Anomalies
sst2_anm_c_1850_2002.grd.gz:
    binary dataset (in the format of SST(360,180,12,153), fortran direct access, little endian, 68MB)
sst2_anm.ctl:
    ctl file for Grads.

5-day Climatologies
sst_clm1pnt_1950_2002.grd.gz:
    binary dataset (in the format of SST(360,180,73), fortran direct access, little endian, 7MB)
sst_clm1pnt_1950_2002.ctl:
    ctl file for Grads.

How to read with Fortran or Matlab.

If you are not familiar with the Grads, you can easily read the data with fortran as follows.
Please use a proper method to read little-endian data, if you use a big-endian machine.

program main
  real:: sst(360,180) !sst(i,j), where i=1 is 0.5E, j=1 is -89.5S with 1 degree resolution. 
  integer:: iyr, mth
  open(11,file='d:\datagrid\myicoads\sst2_anm_c_1850_2002.grd',form='unformatted',access='direct',recl=360*180)   ! intel fortran
! open(11,file='d:\datagrid\myicoads\sst2_anm_c_1850_2002.grd',form='unformatted',access='direct',recl=360*180*4) ! most of other fortran
  do iyr=1850,2002
    do mth=1,12
      nrec=mth+(iyr-1850)*12
      read(11,rec=nrec) sst
    enddo
  enddo
  close(11)
  open(21,file='d:\tmp\x.grd',form='binary')
  write(21) sst
  close(21)
end program 

If you prefer Matlab, please try the following.

fid=fopen('d:\datagrid\myicoads\sst2_anm_c_1850_2002.grd','rb','ieee-le');
undef=-1.e+10;
for yr=1850:2002
  for mth=1:12
%   sst(i,j), where i=1 is 0.5E, j=1 is -89.5S with 1 degree resolution. 
    sst=fread(fid,[360,180],'float32');
    sst(sst==undef)=NaN;
  end
end
contour(sst');

References:
Minobe S. and A. Maeda, 2005: A 1^o SST dataset compiled from ICOADS from 1850 to 2002 and Northern Hemisphere frontal variability. International Journal of Climatology. 25, 881-894. (PDF)

Folland, C. K. and D. E. Parker, 1995: Correction of instrumental biases in historical sea surface temperature data. Q. J. R. Meteorol. Soc., 121, 319-367.