David works on the terrestrial carbon cycle, with a particular interest in improving parametrisations by data assimilation.
Areas of expertise:
terrestrial carbon cycle;
the JULES model;
linearisation of complex models.
David is a scientist working on the terrestrial carbon cycle. He uses the JULES model to simulate the carbon cycle, and uses data from eddy-covariance systems as targets to be hit by JULES' simulations. Some of the parameters are adjusted by an automatic procedure — data assimilation — in which a least-squares measure of the mismatch between simulated and measured time series is minimised. There is scope for large speed-ups in the data assimilation scheme, by producing an adjoint of the JULES code. Since JULES is a large and complex Fortran code, it is very difficult, or practically impossible, to produce adjoint code by hand. Therefore, David is working with the University of Exeter and the FastOpt company in Germany, to produce adjoint code automatically by use of an external adjoint-generating program.
David has a BSc in Physics from the University of York, and a PhD in Plasma Physics from Imperial College, London. His post-doctoral work began at ESSC, University of Reading, on the simulation and measurement of the directional reflectance of vegetation. He followed this with theoretical and practical research on passive microwave remote sensing of soil moisture, in preparation for the MIRAS mission, which is now in orbit aboard the dedicated SMOS satellite.