The new NOAA operational global sea surface temperature (SST) analysis uses 7 days of in situ (ship and buoy) and satellite SST data, produced weekly and daily using optimum interpolation (OI) on a 1° grid. The OI technique requires data and analysis error statistics, which show that SST data errors from ships are twice as large as those from buoys or satellites. The average e-folding spatial error scales are 850 km in the zonal direction and 615 km in the meridional direction. A preliminary step corrects satellite biases relative to in situ data using Poisson's equation. This correction is important, as demonstrated by data following the 1991 eruptions of Mt. Pinatubo. The OI analysis uses in situ and bias-corrected satellite data from 1985 to present. The SST data sources include in situ data from the NMC file of surface marine observations and satellite data from the AVHRR on NOAA satellites. In situ data are sparse near the ice edge, where sea ice information is used. The OI SST analysis is produced daily and weekly on a 1° grid. The analysis uses quality control procedures to eliminate unlikely observations and ensures data accuracy. The OI method uses a least-squares minimization to determine weights, which are used to compute the analysis increment. The first guess is the analysis from the previous week, which is more accurate than climatology for SST forecasts. The OI statistics are computed using data increments and include correlated and uncorrelated error parts. The correlation functions are determined using a negative-squared exponential model. The statistics are used to compute the OI analysis, which is smoother than the blended product. The OI analysis has higher resolution and better represents SST patterns, especially in the tropics and high latitudes. The OI analysis also includes a bias correction step using the Poisson technique to adjust for satellite biases. This correction is important for satellite data affected by stratospheric aerosols, such as those from Mt. Pinatubo. The OI analysis has been shown to be more accurate than previous methods, with a lower error estimate and better resolution. The analysis is sensitive to error statistics, and the new statistics improve the overall fit of the data to the analysis. The OI analysis is used operationally at the NMC and provides a more accurate and detailed SST analysis than previous methods.The new NOAA operational global sea surface temperature (SST) analysis uses 7 days of in situ (ship and buoy) and satellite SST data, produced weekly and daily using optimum interpolation (OI) on a 1° grid. The OI technique requires data and analysis error statistics, which show that SST data errors from ships are twice as large as those from buoys or satellites. The average e-folding spatial error scales are 850 km in the zonal direction and 615 km in the meridional direction. A preliminary step corrects satellite biases relative to in situ data using Poisson's equation. This correction is important, as demonstrated by data following the 1991 eruptions of Mt. Pinatubo. The OI analysis uses in situ and bias-corrected satellite data from 1985 to present. The SST data sources include in situ data from the NMC file of surface marine observations and satellite data from the AVHRR on NOAA satellites. In situ data are sparse near the ice edge, where sea ice information is used. The OI SST analysis is produced daily and weekly on a 1° grid. The analysis uses quality control procedures to eliminate unlikely observations and ensures data accuracy. The OI method uses a least-squares minimization to determine weights, which are used to compute the analysis increment. The first guess is the analysis from the previous week, which is more accurate than climatology for SST forecasts. The OI statistics are computed using data increments and include correlated and uncorrelated error parts. The correlation functions are determined using a negative-squared exponential model. The statistics are used to compute the OI analysis, which is smoother than the blended product. The OI analysis has higher resolution and better represents SST patterns, especially in the tropics and high latitudes. The OI analysis also includes a bias correction step using the Poisson technique to adjust for satellite biases. This correction is important for satellite data affected by stratospheric aerosols, such as those from Mt. Pinatubo. The OI analysis has been shown to be more accurate than previous methods, with a lower error estimate and better resolution. The analysis is sensitive to error statistics, and the new statistics improve the overall fit of the data to the analysis. The OI analysis is used operationally at the NMC and provides a more accurate and detailed SST analysis than previous methods.