Introduction

Introduction


The Atmospheric Model Intercomparison Project (AMIP) is an international effort to simulate the global climate of the period 1979-1988 using certain common boundary conditions and radiative forcings (cf. Gates 1992) [1]. Scientific direction for the AMIP is provided by the Working Group on Numerical Experimentation (WGNE) on behalf of the World Climate Research Programme (WCRP). Detailed coordination of this intercomparison project is the responsibility of the Program for Climate Model Diagnosis and Intercomparison (PCMDI), which is funded by the U.S. Department of Energy (DOE). The AMIP is also coordinated with the DOE Computer Hardware, Advanced Mathematics and Model Physics (CHAMMP) Program (cf. Bader et al. 1992) [2]. In addition, the DOE has provided substantial funding for the execution of AMIP computer simulations at the National Energy Research Supercomputer Center (NERSC) of the Lawrence Livermore National Laboratory (LLNL), making possible the involvement of many modeling groups that would otherwise lack the necessary resources. To date, some 30 groups representing most of the world's global modeling centers are participating in the AMIP.

The boundary conditions common to all the AMIP simulations consist of 120 monthly observed distributions of sea surface temperatures and sea ice extents obtained using available in-situ, climatological, and satellite data. These data were prepared by the Climate Analysis Center of the National Oceanic and Atmospheric Administration (NOAA) in cooperation with the Center for Ocean-Land-Atmosphere Interactions at the University of Maryland. (The PCMDI subsequently disseminated an edited version of these data to each AMIP modeling group.) There are also identical specifications of the solar constant (1365 W/(m^2) and atmospheric carbon dioxide concentration (a uniform 345 ppm) in the AMIP simulations.

In addition, the AMIP effort involves the generation of common monthly mean diagnostics in a standard format, and validation against observational data for the period 1979 to 1988. Diagnosis and validation of the AMIP model simulations is the special responsibility of 26 WGNE-approved diagnostic subprojects, which are focused on the analysis of the simulations of particular physical and dynamical processes or of climatic features of particular regions. Some of these investigations are being coordinated with other WCRP climate initiatives.

A common goal of the diagnostic subprojects is to assess the phenomenological effects of applying diverse numerical schemes and physical parameterizations that are represented in the collection of AMIP models. Given the nonlinear behavior of global atmospheric models, the attribution of particular details of the AMIP simulations to these properties is very difficult and requires, as a minimum prerequisite, an accurate and comprehensive description of the numerics, dynamics, and physics of the models.

The present summary documentation of the AMIP models is written principally to serve this need of the diagnostic subprojects. This document centralizes information on the main features of the AMIP models and expresses this according to a common framework. It is not intended, however, to substitute for existing documentation of these models, which is liberally cited throughout.

The sections of this summary documentation are organized in the order of progressive complexity of information content. That is, an overview of the major differences in the AMIP models' features may be obtained first by perusing the Summary Tables, and further information may then be gleaned from the respective model Summary Reports, which make up the bulk of the document. Additional details on model algorithms and parameterizations can be obtained by consulting the references associated with each model report, or in the comprehensive bibliography.

Return to Main Document Directory

Return to AMIP home page


Last update July 8, 1996. For further information, contact: Tom Phillips (phillips@tworks.llnl.gov)

LLNL Disclaimers

UCRL-ID-116384