Single Precision in the Dynamical Core of a Nonhydrostatic Global Atmospheric Model: Evaluation Using a Baroclinic Wave Test Case
Open Access
- 30 January 2018
- journal article
- research article
- Published by American Meteorological Society in Monthly Weather Review
- Vol. 146 (2), 409-416
- https://doi.org/10.1175/mwr-d-17-0257.1
Abstract
Reducing the computational cost of weather and climate simulations would lower electric energy consumption. From the standpoint of reducing costs, the use of reduced precision arithmetic has become an active area of research. Here the impact of using single-precision arithmetic on simulation accuracy is examined by conducting Jablonowski and Williamson’s baroclinic wave tests using the dynamical core of a global fully compressible nonhydrostatic model. The model employs a finite-volume method discretized on an icosahedral grid system and its mesh size is set to 220, 56, 14, and 3.5 km. When double-precision arithmetic is fully replaced by single-precision arithmetic, a spurious wavenumber-5 structure becomes dominant in both hemispheres, rather than the expected baroclinic wave growth only in the Northern Hemisphere. It was found that this spurious wave growth comes from errors in the calculation of gridcell geometrics. Therefore, an additional simulation was conducted using double precision for calculations that only need to be performed for model setup, including calculation of gridcell geometrics, and single precision everywhere else, meaning that all calculations performed each time step used single precision. In this case, the model successfully simulated the growth of the baroclinic wave with only small errors and a 46% reduction in runtime. These results suggest that the use of single-precision arithmetic will allow significant reduction of computational costs in next-generation weather and climate simulations using a fully compressible nonhydrostatic global model with the finite-volume method.Keywords
Funding Information
- Ministry of Education, Culture, Sports, Science and Technology (FLAGSHIP2020)
This publication has 22 references indexed in Scilit:
- Parallelization and Performance of the NIM Weather Model on CPU, GPU, and MIC ProcessorsBulletin of the American Meteorological Society, 2017
- On the use of programmable hardware and reduced numerical precision in earth‐system modelingJournal of Advances in Modeling Earth Systems, 2015
- Benchmark Tests for Numerical Weather Forecasts on Inexact HardwareMonthly Weather Review, 2014
- Scalable rank-mapping algorithm for an icosahedral grid system on the massive parallel computer with a 3-D torus networkParallel Computing, 2014
- The use of imprecise processing to improve accuracy in weather & climate predictionJournal of Computational Physics, 2014
- Predictability and Prediction Skill of the MJO in Two Operational Forecasting SystemsJournal of Climate, 2014
- A Finite-Volume Icosahedral Shallow-Water Model on a Local CoordinateMonthly Weather Review, 2009
- A baroclinic instability test case for atmospheric model dynamical coresQuarterly Journal of the Royal Meteorological Society, 2006
- A Proposal for the Intercomparison of the Dynamical Cores of Atmospheric General Circulation ModelsBulletin of the American Meteorological Society, 1994
- The Simulation of Three-Dimensional Convective Storm DynamicsJournal of the Atmospheric Sciences, 1978