Common Infrastructure for Modelling the Earth (CIME)

Common Infrastructure for Modelling the Earth (CIME)
Name	Common Infrastructure for Modelling the Earth
Abbreviation	CIME
Developers	National Center for Atmospheric Research; National Oceanic and Atmospheric Administration; Lawrence Livermore National Laboratory
Initial release	2007
Programming languages	Fortran; Python; Bash
Platform	Linux; macOS
License	open-source

Common Infrastructure for Modelling the Earth (CIME) is a software framework used to configure, build, and run coupled atmosphere, ocean, land, and sea-ice models in large-scale Earth system modelling projects. It supports model integration and workflow orchestration across research groups at institutions such as National Center for Atmospheric Research, NASA, NOAA and national laboratories like Lawrence Livermore National Laboratory and Los Alamos National Laboratory, facilitating reproducible experiments for projects related to Coupled Model Intercomparison Project, CMIP6 and US Department of Energy initiatives.

Overview

CIME provides a standardized build and run system that unifies model components from projects including Community Earth System Model, Energy Exascale Earth System Model, GFDL CM4 and UK Met Office Unified Model while interoperating with workflow engines used by European Centre for Medium-Range Weather Forecasts, NERSC and Argonne National Laboratory. The framework automates tasks such as case creation, dependency management, component coupling and resource allocation on systems from commodity clusters to supercomputers like Oak Ridge Leadership Computing Facility and Summit (supercomputer). CIME’s interfaces are implemented in shell and Python to integrate with build tools and job schedulers such as SLURM, PBS, and LSF.

History and Development

CIME originated from collaborative efforts led by developers at National Center for Atmospheric Research and Los Alamos National Laboratory to consolidate disparate scripts used in projects like CESM1 and earlier Community Climate System Model versions. Subsequent development drew contributions from teams at NOAA Geophysical Fluid Dynamics Laboratory, Princeton University, and Lawrence Livermore National Laboratory to support multi-component coupling needs identified during Coupled Model Intercomparison Project Phase 5 and the planning for CMIP6. Major milestones include migration to modular Python interfaces inspired by practices at NASA Goddard Space Flight Center and adaptation for exascale preparation in coordination with DOE workshops and XSEDE partner sites.

Architecture and Components

CIME’s architecture separates platform abstraction, case management, build system, and run-time coupling glue, integrating components such as the Community Atmosphere Model, Parallel Ocean Program, Community Land Model, and Community Ice CodE. Key modules include a case creation utility that reads XML configuration files, a build system that generates component executables, and a run script that orchestrates coupling with flux exchanges conforming to standards promoted by Earth System Modeling Framework and ESMF. The software relies on toolchains including compilers from Intel Corporation and GNU Project toolchain components, MPI libraries like Open MPI and MPICH, and external libraries such as NetCDF and HDF5.

Workflow and Usage

Users create a case referencing machine and component configurations, then use CIME commands to build and submit experiments to queuing systems at facilities like National Energy Research Scientific Computing Center and Oak Ridge National Laboratory. Typical workflows connect model output processing to analysis packages developed at University Corporation for Atmospheric Research and visualization tools used by Visualization Toolkit and ParaView. CIME supports ensemble workflows for projects hosted by World Climate Research Programme and automated verification suites aligned with NASA Earth Exchange practices, enabling reproducible experiments across collaborative centers such as Scripps Institution of Oceanography and Lamont–Doherty Earth Observatory.

Governance and Community

Development and governance follow an open collaborative model with steering from stakeholders at National Center for Atmospheric Research, NOAA, DOE laboratories, and universities including University of Washington and University of Colorado Boulder. Contributions, issue tracking, and releases are coordinated via platforms used by many scientific software communities and involve working groups that mirror organizational structures found in Coupled Model Intercomparison Project task teams and World Climate Research Programme panels. Training and outreach occur through tutorials at conferences such as American Geophysical Union and Ocean Sciences Meeting, and community support is provided through mailing lists and repositories shared among partners like GitHub and institutional servers at Argonne National Laboratory.

Applications and Case Studies

CIME underpins production simulations for climate assessment reports prepared by Intergovernmental Panel on Climate Change authors, regional downscaling efforts coordinated with North American Regional Climate Change Assessment Program, and targeted studies by researchers at NOAA Geophysical Fluid Dynamics Laboratory and University of Oxford. Case studies include coupled hindcasts for attribution studies cited in IPCC Fifth Assessment Report style analyses, high-resolution regional climate projections for California Energy Commission planning, and biogeochemical coupling experiments used in ecosystem assessments by National Oceanic and Atmospheric Administration scientists.

Licensing and Distribution

CIME is distributed under open-source licensing terms aligned with common practices used by projects like Community Earth System Model and hosted on collaborative code platforms adopted by institutions such as National Center for Atmospheric Research and Lawrence Livermore National Laboratory. Binary builds and modules are packaged for HPC centers including NERSC and Oak Ridge Leadership Computing Facility while source releases are versioned to support citation and reproducibility in publications submitted to journals like Geophysical Research Letters and Journal of Climate.

Category:Earth system modeling