The meeting will start with a few plenary talks and a poster session to highlight the state-of-the-art in computational scattering science. Smaller workgroups will provide forums for discussing where the individual fields are presently going, and where they should go in the future. Through the workgroups, attendees will contribute to a strategic plan to design, develop, deploy and sustain software workflows to advance scattering science with x-rays and neutrons.
For experimental scattering science, a scientific workflow is a sequence of steps that takes data from the lab to publication. Computation is essential at every step, for example for reducing experimental data, developing models of structure or dynamics, and for finding interrelationships or inconsistencies between results from computational science and experiment.
The two documents available on this website describe in some detail the opportunities for better incorporation of modern computational science into different fields of scattering science. These fields span from the life sciences through physics, centered around the materials studied by scattering experiments with x-rays, neutrons, and electrons. Some broad themes involve the use of modern codes for simulations of materials structure and dynamics to predict the scattering from materials, perhaps in full simulations that include the instruments themselves. Examples of obvious computational science packages of today are ab-initio electronic structure packages, molecular dynamics packages, but there are needs for more sophisticated tools to handle magnetism, correlated electron phenomena, and the excited electronic states that occur in spectroscopic measurements on materials. Some of these tools are on the near horizon; others are still emerging.
Today computational scientific workflows consist of islands of standalone packages with focused purpose. For many computational scattering science problems, the island structure is not complete, bridges between the islands are not always in good repair, and some are missing. The importance of some islands is not yet fully appreciated. On the other hand, bridges have been constructed to controversial islands. What should this infrastructure include?
Most software components of today are not easily separated for inclusion into new workflows. A greater range of workflow configurations can be built with more modular software, but greater flexibility requires more user sophistication. How is this balance best determined? What investments in software flexibility and user support offer the best value?
It is likely that new software workflows will continue to be developed by different groups under different styles of management and interactions between scientists and software developers. The workshop will address how this is best accomplished for the broader scattering community. Can community demand be organized to evaluate and build new software workflows, rather than relying on the cottage industries that exist today?
Future software development will involve separate groups working at different locations. How best can these efforts be coordinated? Is there an appropriate governance for such a distributed activity?
The first day is for a few plenary talks and posters that describe the present status and plans for computational scattering science.
In the next two days, each attendee will participate in approximately three workgroups. At the discretion of the topic leader, there will be brief presentations and extended discussion during the first day. On the second day, the workgroups will write a short report with proposals for where the field could be in 5 years, and what paths are available to get there.
In unscheduled time, we hope that interest groups will be organized with advocacy and collaboration for the development of specific workflows.