CSE Software Ecosystems: Critical Instruments of Scientific Discovery

Software is an essential product of CSE research when complex models of reality are cast into algorithms; moreover, the development of efficient, robust, and sustainable software is at the core of CSE. While the community is beginning to embrace the fundamental role of open-source software ecosystems to support CSE collaboration and enable advances in scientific and engineering understanding, much work remains to overcome challenges in software sustainability and productivity. Difficulties arise from the confluence of disruptive changes in computing architectures, new opportunities for greatly improved simulation capabilities, and demand for greater scientific reproducibility. New architectures require fundamental algorithm and software refactoring, while at the same time enabling new multiscale and multiphysics modeling, simulation, and analysis.

This presentation will introduce activities under way throughout the community to address a variety of technical and social issues in scientific software. We will also highlight multipronged work in the IDEAS Scientific Software Productivity Project to distill and promote best practices in software engineering and productivity for CSE, develop the xSDK as a foundation of a community CSE software ecosystem, use these best practices and software to achieve science advances (e.g., in subsurface flow with hydrological and biogeochemical recycling), and engage the community in collaborative contributions.