1Ministry of Education Key Laboratory for Earth System Modeling, Department of Earth System Science, Tsinghua University, Beijing, China
2Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), China
3CMA Earth System Modeling and Prediction Centre (CEMC), China
4First Institute of Oceanography, and Key Laboratory of Marine Science and Numerical Modeling, Ministry of Natural Resources, Qingdao, China
5Unit No. 91001 of PLA, Beijing, China
6State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
1Ministry of Education Key Laboratory for Earth System Modeling, Department of Earth System Science, Tsinghua University, Beijing, China
2Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), China
3CMA Earth System Modeling and Prediction Centre (CEMC), China
4First Institute of Oceanography, and Key Laboratory of Marine Science and Numerical Modeling, Ministry of Natural Resources, Qingdao, China
5Unit No. 91001 of PLA, Beijing, China
6State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China
Received: 12 Mar 2022 – Discussion started: 04 May 2022
Abstract. As Earth system modeling develops ever finer grid resolutions, the inputting and outputting (I/O) of the increasingly large data fields becomes a processing bottleneck. Many models developed in China, as well as the Community Coupler (C-Coupler), do not fully benefit from existing parallel I/O supports. This paper reports the design and implementation of a Common parallel Input/Output Framework based on C-Coupler2.0 (CIOFC1.0). Parallelization by CIOFC1.0 can accelerate the I/O of large data fields. The framework also allows convenient specification by users of the I/O settings; e.g., the data fields for I/O, the time series of the data files for I/O, and the data grids in the files. The framework can also adaptively input data fields from a time-series dataset during model integration, automatically interpolate data when necessary, and output fields either periodically or irregularly. CIOFC1.0 demonstrates the cooperative development of an I/O framework and coupler, and thus enables convenient and simultaneous use of a coupler and an I/O framework.
In this paper, we propose a new common, flexible and efficient parallel I/O framework for Earth system modeling based on C-Coupler2.0. CIOFC1.0 can handle data I/O in parallel, and provides a configuration file format that enables users to conveniently change the I/O configurations. It can automatically make grid and time interpolation, output data with an aperiodical time series and accelerate data I/O when the field size is large.
In this paper, we propose a new common, flexible and efficient parallel I/O framework for Earth...