Establish governance rules for new ESiWACE developments (MS3)

Effective governance necessitates intense exchange within and beyond the community. We have noted in the ESiWACE proposal (Sec. 1.3.3) that “networking activities depend on the community’s maturity in a given area. In some areas we can build on previous activities around established shared software especially from IS-ENES2. This includes the evaluation of coupling technologies and the support of the OASIS coupler, the development of I/O strategies, the development and support of the XIOS I/O server and the Cylc workflow engine, common analyses of model codes such as NEMO and EC-Earth. In this case networking for existing community software will foster community co-operation based on an agreed set of benchmarks and will drive continued research towards improvement of these tool through engagement with the existing user community. Dissemination of community software will be enhanced by improved user support and training as well as fastest possible integration of new software versions.
In other areas, the community’s maturity is lower and will focus on networking for new community software with activities supporting to share specialist knowledge that exists but is not yet available across the community. An example is the development of suitable software stacks for HPC systems.”
Central goal of ESiWACE is to improve efficiency and productivity for weather and climate simulations in view of upcoming exa-scale computing and data handling systems. New software developments arise in this context, besides supporting and improving existing software solutions.
During the ESiWACE General Assembly meeting in July 2016, the consortium agreed to focus on the implementation and analysis of demonstrators for very high resolution Earth System Models in order to identify gaps between current and expected exa-scale weather and climate simulations and, correspondingly, prioritise developments. These demonstrators shall yield insights into computability of weather and climate codes in the era of exa-scale computing. Hence, they shall provide evidence to develop a roadmap for driving climate and weather simulation towards exa-scale. The interest in the demonstrator development is significant, which has recently been confirmed by a survey on community software development: 80% of non-ESiWACE members (scientific and industrial institutions) who participated in this survey stated that they want to be (regularly) informed on the ESIWACE demonstrator developments.