Gridded temperature anomalies for monthly mean temperature are provided from the GHCN v3 monthly temperature data set (grid-mntp-1880-current-v3.y.z.dat.gz). In most cases users will want to access the most recent version.

Each month of data consists of 2592 gridded data points
produced on a 5 X 5 degree basis for the entire globe (72 longitude X 36 latitude grid boxes).
   
Gridded data for every month from January 1880 to the most recent month are available. The data are temperature anomalies in hundredths of a degree Celcius.  Each gridded value was multiplied by 100 and written to file as an integer.  Missing values are represented by the value -9999.

The data are formatted by year, month, latitude and longitude.  There are 72 longitude grid values per line. Longitude values are written from 180 W to 180 E, and latitude values from 90 N to 90 S.  Data for each month is preceded by a label containing the month and year of the gridded data.

grid-mntp-1880-current-v3.y.z.dat
     for year = begyr to endyr
      for month = 1 to 72
        format(2i6) month,year
        for ylat = 1 to 36 (85-90N,80-85N,...,80-85S,85-90S)
          format(72i5) 180-175W,175-170W,...,170-175E,175-180E
   

Each file has been compressed using 'gzip'.  It can be uncompressed with 'WinZip' for those using Windows or with 'gzip' from most UNIX platforms.  
   
Temperature anomalies are calculated with respect to the period 1961 - 1990 using the Climate Anomaly Method. The gridded data sets are produced to provide the most accurate time series possible. However, this required that we treat months and grid boxes independently through time.  The use of these data is most appropriate for analyzing the change in temperature within a particular grid box, or set of grid boxes, over a span of years.  If one is more interested in analyzing temperature changes within individual years, e.g., the change in temperature between February and March, 1908, or between two regions in 1908, we recommend that the GHCN station data be used
directly.