Articles | Volume 5, issue 2
Geosci. Model Dev., 5, 449–455, 2012
Geosci. Model Dev., 5, 449–455, 2012

Development and technical paper 03 Apr 2012

Development and technical paper | 03 Apr 2012

"Gtool5": a Fortran90 library of input/output interfaces for self-descriptive multi-dimensional numerical data

M. Ishiwatari1, E. Toyoda2, Y. Morikawa3, S. Takehiro4, Y. Sasaki5, S. Nishizawa6, M. Odaka1, N. Otobe7, Y. O. Takahashi6, K. Nakajima8, T. Horinouchi9, M. Shiotani10, Y.-Y. Hayashi6, and Gtool development group11 M. Ishiwatari et al.
  • 1Hokkaido University, Graduate school of Science, Sapporo, Japan
  • 2Japan Meteorological Agency, Tokyo, Japan
  • 3Parmy, Inc., Japan
  • 4Kyoto University, Research Institute for Mathematical Sciences, Kyoto, Japan
  • 5Kyoto University, Department of Mathematics, Kyoto, Japan
  • 6Kobe University, Graduate school of Sciences, Kobe, Japan
  • 7Fukuoka University, Faculty of Science, Fukuoka, Japan
  • 8Kyushu University, Faculty of Science, Fukuoka, Japan
  • 9Hokkaido University, Graduate school of Environmental Earth Science, Sapporo, Japan
  • 10Kyoto University, Research Institute for Sustainable Humanosphere, Kyoto, Japan
  • 11GFD Dennou Club, Japan

Abstract. A Fortran90 input/output library, "gtool5", is developed for use with numerical simulation models in the fields of Earth and planetary sciences. The use of this library will simplify implementation of input/output operations into program code in a consolidated form independent of the size and complexity of the software and data. The library also enables simple specification of the metadata needed for post-processing and visualization of the data. These aspects improve the readability of simulation code, which facilitates the simultaneous performance of multiple numerical experiments with different software and efficiency in examining and comparing the numerical results. The library is expected to provide a common software platform to reinforce research on, for instance, the atmosphere and ocean, where a close combination of multiple simulation models with a wide variety of complexity of physics implementations from massive climate models to simple geophysical fluid dynamics models is required.