One upon a time I was an astrophysicist. Then I realised that my motivation didn’t come from the stars (interesting as they are), but instead from solving tough scientific problems using sustainable software.

I’m currently leading a small team of Scientific Software Engineers as part of a greater effort to update the Met Office’s ensemble data assimilation system. I’m keen on promoting diversity of thought as well as agile practices such as pair programming and test-driven development.

Beyond line management, I have over fifteen years of experience in writing massively parallel software for high performance computing clusters. I’m particularly adept in figuring out solutions for tricky numerical geometry problems. I’m an expert C++ developer with an unhealthy fascination towards template metaprogramming. I’ve also written my fair share of Python and Fortran and I’d say that I’m Rust-curious.

Science-wise, I’m working in the field of data assimilation for numerical weather prediction. This Bayesian method combines an a priori ensemble of short-range Earth-system forecasts with vast quantities of observations, resulting in an a posteriori distribution of model states. These represent the best understanding of the “true” state of the system, and are used to initialise longer range ensemble weather forecasts.