NOAA's Merged Land-Ocean Surface Temperature Analysis (MLOST) combines land surface air temperatures primarily from the Global Historical Climatology Network, Monthly (GHCN-M) version 3 with SSTs of the ERSSTv3b analysis into a comprehensive global surface temperature data set spanning 1880 to the present at monthly resolution, on a 5x5 degree latitude-longitude grid. As such, it is one of the main data sets used to monitor global and regional temperature variability and trends. Like GISTEMP, interpolation of the station records is performed to provide broad spatial coverage, however, the methodology is unique. Land and ocean domains are treated separately. For both domains, high and low-frequency variations are reconstructed separately using empirical orthogonal functions, and these signals are combined and blended with the original records. The final step involves combining the land and ocean data sets into one; over coastal regions land and ocean values are weighted in proportion to the amount of land or ocean in the 5x5 degree box. Data-sparse areas are masked to prevent reliance on the highly smoothed reconstruction estimates. Sea ice areas are also set to missing.