Accurate and efficient parameterizations of the various physical processes that are necessary to simulate Earth’s climate are critical for the ACME Earth System Model (ESM) to achieve its key scientific goals. The complexity of radiative transfer, its importance to the exchange of energy in the climate system, and its high computational cost make this component especially important in this context. Previous work by the proposing team at AER led to the development of the radiation code, RRTMG, which has been widely accepted by the global modeling community as a fast and accurate advancement over the previous generation of radiation codes. It has been in use in the NCAR CESM for many years, and it is part of the current version of the ACME ESM. However, its computational cost in CESM remains high relative to other components in part due to its complexity and to its inefficient use of modern optimization strategies, and it will be advantageous to address this limitation for the code’s application to the ESM.
Under other funding, the Investigators of this project are leading an effort to develop a high-performance broadband radiation code, called RRTMGP, which is a completely restructured version of RRTMG that will take advantage of modern capabilities to enhance its performance while retaining the strengths and accuracy of the original code. Designed to perform over a range of computer architectures, RRTMGP makes extensive use of Fortran 2003 features to improve both its efficiency and its use of memory. Development of RRTMGP for highly parallel Central Processing Unit (CPU) platforms will be completed before the end of 2016, and it is expected that this code will be adopted in the next generation of CESM. In collaboration with the ACME development team during this project, RRTMGP and its associated interfacing will be adapted as necessary to couple it efficiently and effectively to the related ESM physics components, most notably the aerosol and cloud specification modules. This effort will focus on enhancing optimization on the limited number of emerging computing systems on which the ESM is expected be used, including Many Integrated Core (MIC) architectures and Graphics Processing Unit (GPU) hardware. We will also develop additional capabilities for RRTMGP that ACME scientists identify as important for the planned applications of the model. This project will optimize a key physical component of the ACME ESM with the goal of directly supporting ACME’s overarching global modeling objectives.