AMIP Sea Surface Temperature and Sea Ice:
Observational and Boundary Condition Data Sets

The AMIP SST and sea ice boundary conditions are available at this site. Two distinctly different data sets have been prepared. The first contains observed monthly mean SSTs, which area are provided and updated following a procedure described by Hurrell et al. (2008). This dataset, which at this site became available in 2007, supercedes the original AMIP II observational data set described here. There are differences between the earlier dataset and the present one which are generally small but not entirely negligible.

The second data set is the AMIP SST and sea ice boundary conditions derived from the observational data set as described in this web document. This data should be used in the AMIP simulations. The primary difference between the boundary condition data set and the observational data set is that adjustments have been made to insure that when daily values are linearly interpolated between the monthly values contained in the data set (i.e., when used as boundary conditions for an AGCM), the monthly average of the observations is preserved. Monthly averages are not preserved when daily values are obtained by linearly intepolated between monthly mean observations (as in AMIP I). When the boundary condition data (as opposed to the observed monthly means) are used for an AGCM simulation, it has been shown that in addition to preserving the observational monthly means, the seasonal cycle and the interannual variance are more accurately represented than the method used in AMIP I.

Download AMIP Boundary Condition Data at 1 by 1 Degree Resolution

The proper citation for the boundary condition data is the following report: Taylor, K.E., D. Williamson and F. Zwiers, 2000: “The sea surface temperature and sea ice concentration boundary conditions for AMIP II simulations” PCMDI Report 60, Program for Climate Model Diagnosis and Intercomparison, Lawrence Livermore National Laboratory, 25 pp available as a PDF (For further descriptive details, see here). Instructions on how to correctly prescribe the boundary conditions in AMIP simulations are contained in those documents. The data sets will be periodically updated to near-present (every 6 months, or so).

Note that earlier versions (prior to 2007) of the observational dataset on which the boundary conditions are based were prepared following a different procedure described here. There are differences between the earlier dataset and the present one which are generally small but not entirely negligible.

Document version: 21 September 2023