Dr Ian C. Rutt
 |
Lecturer in Glaciology Department of Geography
Tel. 01792 602685
Email: i.c.rutt@swansea.ac.uk |
Research Interests
Ice sheet modelling
I am a glacial modeller, with particular experience of coupling ice sheet models to climate models.
I have played a central role in the development of the Glimmer community ice sheet model. Glimmer is a regional ice model that can be flexibly coupled to a GCM, through a highly-configurable interface. Glimmer is based on a stand-alone model written by Tony Payne, though substantial development work has been done to bring the model to its current state. This work has been done in collaboration with Magnus Hagdorn and Nick Hulton at the University of Edinburgh.
Glimmer has a number of attractive and novel features:
- May be run simultaneously over any number of regions of the globe without code changes or recompilation;
- Has multiple schemes available for different elements of the model (mass-balance, dynamics, etc);
- Modular structure - may be run in stand-alone mode, with custom climate drivers, or coupled to a GCM;
- Simple and flexible configuration files;
- Output in standard netCDF format, for analysis with CDAT, Ferret, Matlab, etc;
- Any number of output files may be specified;
- Written throughout in Fortran 95;
- Free software, licensed under the GNU GPL.
Glimmer has a growing community of users in the UK, the US, China and Europe. Development is ongoing, with regular code releases available for download here.
Other interests
In connection with my work in glacial modelling and PhD work in atmospheric modelling, I have developed a strong interest in the application of modern computer-science techniques to scientific programming. Numerical models form the foundations on which much of the geosciences are built, but while the numerical and mathematical approximations used in these models are the subject of scrutiny and discussion, the way that code is developed is often given less thought.
Historically, Fortran has been the pre-eminent language for numerical work, and this is still true. However, much scientific code is still written in Fortran 77, using 30 year-old programming techniques. F77 is quite adequate for small projects, but the structural aspects of the language were never designed for the kind of large-scale models which we develop now. Particularly significant limitations are:
- Lack of dynamic memory management facility
- COMMON blocks and EQUIVALENCE - great potential for error
- No checking of subroutine interfaces on compilation
- No standard connectivity with operating system or other languages (e.g. C)
- Limited extensibility; no object-orientation
Some of these characteristics encourage poor programming practices, leading to 'spaghetti code' that is difficult to debug or reuse. Fortran 90/95 addresses some of these deficiencies, though the provision for object-orientation is fairly limited; a thorough discussion of object-orientation in f90/95 can be found on this website, thanks to work by Viktor Decyk, Charles Norton and Boleslaw Szymanski.
However, even given additional language features, the most appropriate design strategy for a numerical model remains a matter of judgement, and more than a little alchemy. This is something I try and explore in my modelling work.
In addition to languages and programming techniques, I am interested in more general issues in the practice of model development, especially code provenance, version control and automated testing.
Teaching
I will take on teaching responsibilities in the autumn term, 2007.
Publications
Journal articles
Bougamont M, Bamber JL, Ridley JK, Gladstone RM, Greuell W, Hanna E, Payne AJ, and Rutt I (2007) Impact of model physics on estimating the surface mass balance of the Greenland ice sheet. Geophys Res Lett 34, L17501, doi:10.1029/2007GL030700.
Lenton TM, Marsh R, Price AR, Lunt DJ, Aksenov Y, Annan JD, Cooper-Chadwick T, Cox SJ, Edwards NR, Goswami S, Hargreaves JC, Harris PP, Jiao Z, Livina VN, Payne AJ, Rutt IC, Shepherd JG, Valdes PJ, Williams G, Williamson MS, Yool A (2007) Effects of atmospheric dynamics and ocean resolution on bi-stability of the thermohaline circulation examined using the Grid ENabled Integrated Earth system modelling (GENIE) framework. Clim Dyn (published online. doi:10.1007/s00382-007-0254-9
Lunt DJ, Valdes PJ, Haywood A, Rutt IC (2007) Closure of the Panama Seaway during the Pliocene: implications for climate and Northern Hemisphere glaciation. Clim Dyn (published online. doi:10.1007/s00382-007-0265-6)
Payne AJ, Holland PR, Shepherd AP, Rutt IC, Jenkins A, Joughin I (2007) Numerical modeling of ocean-ice interactions under Pine Island Bay's ice shelf. J. Geophys. Res., 112, C10019, doi:10.1029/2006JC003733
Rutt IC, Thuburn J, Staniforth A (2006) A variational method for orographic filtering in NWP and climate models. Q. J. R. Meteorol. Soc. 132, 1795-1813
Conference presentations and other publications
Rutt IC, Payne AJ, Lunt DJ, Valdes PJ (2005) Simulation of seasonal snow-cover and glacial inception with GENIE and GLIMMER. Oral presentation given at AGU Fall Meeting, 2005 (Eos Trans. AGU, 86 (52) Fall Meet. Suppl., Abstract C12A-01)
Rutt, IC (2005) Writing software is far from trivial (letter). Physics World 18 (8), 22
Rutt, IC, Payne AJ, Lunt DJ, Valdes PJ (2005) Modelling glacial inception with GENIE (poster). Presented at EGU General Assembly, Vienna 24-29 April 2005
Rutt, IC (2003) Optimal filtering for orography in NWP and climate modelling. PhD thesis, University of Reading.
Rutt, IC (2002) Filtering near grid-scale orography in a shallow water model (poster). Presented at EGS General Assembly, Nice, 21-26 April 2002