Scientific Software Engineer - Earth Observation Data Assimilation

at  Met Office

Job Introduction
We’re looking for an exceptional Scientific Software Engineer - Earth Observation Data Assimilation, to help us make a difference to our planet.
As our Scientific Software Engineer, the job may be suitable for hybrid working, which is where an employee works part of the week in the office and part of the week from home. This is a voluntary, non-contractual arrangement and the location advertised will be your contractual place of work.
Our opportunity is full time, 37 hours per week, but we would also consider applicants wishing to work part time and we will also try our best to consider those intending to work a job share. Our people are at the heart of what we do and we’ll do our best to agree a working pattern that works for everyone.

YOUR WORLD OF EXPERTISE

The Met Office has embarked on an exciting and ambitious programme to redesign its operational forecasting capability for future high-performance computing (HPC) platforms including its Data Assimilation (DA) system; ensuring it will be efficient across the range of possible alternative HPC solutions likely to be available on the 2025 timescale. The new software is being developed within the Joint Effort for Data assimilation Integration (JEDI) framework from the Joint Center for Satellite Data Assimilation (JCSDA).
You will work for one of the scientific software engineering teams in Weather Science responsible for the development and delivery of the new JEDI software. Data assimilation is the process of combining observational data with output from a numerical model to produce a statistically optimal estimate of the initial conditions for weather prediction. The Met Office is at the forefront of Observation processing and DA applications in operational Numerical Weather Prediction (NWP) and it is an exciting time to be working in this area, with significant advances in the science of DA, and increasing complexity of modern DA algorithms through the development of coupled DA across Earth System Model components. Perhaps the greatest challenge for modern DA is the need to consider how these complex DA algorithms will be run efficiently on the potentially radically different supercomputer architectures needed to maintain progress over the next decade.
As part of our approach to growing expertise in DA and JEDI, the successful applicants will also have exciting secondment opportunities to spend time working in Boulder, Colorado, at the JCSDA.
We are looking for talented and creative candidates eager to contribute to the research and development of satellite and/or surface observation assimilation techniques for initialising operational weather forecast models on the next generation of supercomputers. The successful applicants will become members of our research and development team and will apply scientific and technical knowledge to contribute to the success of the Met Office.

ESSENTIAL CRITERIA, SKILLS AND EXPERIENCE:

• Modern C++ development experience (C++11 or later). (Expert by Nature)
• Experience in applying scientific/technical knowledge in innovative ways to deliver robust software. (Keep Evolving)
• Demonstrable ability to be a strong team player with an ability to contribute effectively to scientific group activities. (Better Together)
• Experience of using object-oriented design, including algorithms and data structures. (Expert by Nature)
• Experience with software development for parallel or heterogeneous (e.g., GPU/CPU) computing.
• Evidence of applying and improving best practices in quality assurance, including version control, testing, and reviewing.

To contribute towards the development of technical solution and high-quality C++ code according to strict scientific and operational requirements.
For this effort, we are seeking a confident C++ Developer to join our team of experts and contribute toward the technical development and optimization of a unified, next-generation observation processing and data assimilation system that will take full advantage of the next generation HPC. Contributions to the design and development of the JEDI code base could for example include the extension of current class definitions to improve flexibility, usability, and data exchange mechanism between different components of the data processing pipeline.