The global skills and competency framework for the digital world

Computational science skills in SFIA

Digital technology and computational numeracy are becoming increasingly important to all aspects of life and this is also true for the way science is done. Many thanks to the working group in Australia who have been looking at how SFIA can be adapted to describe the skills needed in this area.

Our increasing reliance on data, sensor networks, numerical modelling, data mining, machine learning and inference and inversion techniques require a different and emerging skillset to develop critical science outcomes for the future.

These disciplines are often collectively referred to as Computational Science or Scientific Computing. The associated skills are in high and growing demand.

For some examples - here are links to science/research topics at Australian Public Service (APS) science organisations...

SFIA 8 working group

SFIA describes several digital skills in use but has not yet incorporated those of Computational Science.

A working group has been defining these computational science skills in the context of the SFIA framework.

By specifying these skills in sufficient detail within SFIA - organisations will be able to map out the skills needs and ultimately be able to attract, grow and retain talent and capability.

Breaking down Computational Science, the following three skills were identified.

  • Scientific modelling
  • Numerical analysis
  • High performance computing

The working group has made good progress and has creating prototype definitions for these skills. The work is ongoing - if you would like to help develop and refine these please get in touch.

Change request #1347 Computational Science skills

Prototyping new or changed SFIA skills is a quick and effective way of testing the requirements and getting feedback.

Definitions of the 3 skills

Scientific modelling The identification of the relevant mathematical principles and scientific theory within an information systems framework, to solve real-world problems. The creation, testing and tuning of scientific models through the application of computing. The validation and interpretation of models implemented in information systems against the reality which models attempt to represent.
Numerical analysis Numerical analysis is the area of mathematics and computer science that creates, analyzes, and implements algorithms for solving numerically the problems of mathematics. Numerical Analysis is concerned with floating point arithmetic and the resulting accumulation of rounding errors as opposed to integer arithmetic which has different considerations. Numerical analysis is required for most applications that implement simulations of physical systems, machine learning, data analytics etc.
High performance computing High-performance computing (HPC) is the ability to process data and perform complex calculations at much higher speeds than what is possible on normal desktop computers. The skills involve knowing when to apply HPC as well the ability to implement software to efficiently use HPC architectures.
The intense application of computing resources to achieve very high volumes of computation, including high speed linear processing, scalable parallel processing and sophisticated data manipulation. The development and application of techniques and algorithms that enable the effective use of advanced computer systems.