AMIGO: An Argo-based Model for Investigation of the Global Ocean physics The AMIGO is a new P.P.Shirshov Institute of Oceanology numerical atlas, providing global maps of ocean variables to support the scientific community with near-real time, globally unified, user friendly, research quality dataset. The model uses Argo data to diagnose monthly one-degree global ocean temperature, salinity, and velocities best agreeing with Argo profiles and Aviso altimetry. Argo profiles variationally interpolated onto a vertical and horizontal grid are combined with the WOA09 climatology for layers below 2000 m and ECMWF winds into an ocean general circulation model (OGCM) to produce three-dimensional temperature, salinity, and velocities fields in the global ocean. The principle of the variational interpolation technique is to minimize the misfit between the interpolated fields defined on a regular grid and the irregularly distributed data, so the optimal solution passes as close to the data as possible. OGCM's data analysis is performed in two steps. At the first step the model velocity field is adjusted to the Argo-based T/S fields via forcing model equations of motion by the corresponding steady state winds and integrating them forward in time until all the fast transient processes are dumped. After that the full model is allowed to evolve freely under the stationary forcing fields at the surface. The integration is terminated after 10-30 days. This time is sufficient to remove fast trends from the solution. On the other hand the solution does not deviate too far from data. Spatial resolution - 1 degree global Vertical resolution - 32 levels from 10 to 5500 m Temporal coverage - since 2005-01-01 Temporal resolution - monthly Files are initially contained in one tar file and compressed using the gzip utility. The names of the tar files conform with the following convention: VVYYYYMM.tar.gz. Where: VV represents the variable using two-letters code (TS or UV) YYYY represents the year (0000 corresponds to 2005-2014 climatology) MM represents the averaging period, two digit code as follows: 00 – annual data; 13 to 16 – seasonal data: season 13 - winter (first three months of the year - Jan-Mar); season 14 – spring (Apr-Jun); season 15 – summer (Jul-Sep); season 16 – autumn (Oct-Dec); 01 to 12 – monthly data (starting with 01 – January, to 12 – December); Konstantin Lebedev KLebedev@ocean.ru