Dr Keith Williams
Keith leads research to evaluate and understand, from a process-based perspective, the cause of errors in the global Met Office Unified Model
Keith is the scientific manager of the Model Evaluation team. The work of this team is focused on evaluating the global Unified Model on time and space scales appropriate to the processes being assessed, identifying processes which are not being simulated correctly, and working with others to address these deficiencies in order to improve our predictions across all time scales. A key aspect of the work is developing novel diagnostic techniques to assess the model against a variety of observational data (satellite data, intense observing sites, re-analyses, etc.).
Keith's personal research is in evaluating cloud processes across timescales. This involves detailed comparison of the model against active and passive satellite data (using satellite simulators in the model to emulate the characteristics of the space-borne instruments), combined with surface-based cloud observations. A variety of diagnostic techniques (initial tendencies, nudging, sensitivity experiments, etc.) are employed to help identify the source of any errors.
Keith obtained his degree in physics and meteorology, and his PhD, from the University of Reading. Upon joining the Met Office in 1998, Keith focused on understanding processes leading to a spread in climate sensitivity between climate models. Most of this work involved understanding the spread in cloud response to climate change between models, and determining confidence in their response through evaluation of relevant processes in the current climate.
In 2008, Keith became the manager of the Seamless Model Assessment team which began the work of drawing together model assessment techniques from the weather and climate communities. Following a re-structuring of the Met Office science areas in 2010 to promote more seamless working, Keith began leading the newly formed Model Evaluation team.
- Co-chair of the World Meteorological Organization's Working Group on Numerical Experimentation (WGNE). This group has the responsibility of fostering the development of atmospheric circulation models for use in weather, climate, water and environmental prediction on all time scales and diagnosing and resolving shortcomings in these models.
- Chair of the WGNE-WGCM Transpose-AMIP project. This project aims to run climate models in 'weather forecast mode', by initialising them with re-analyses and running five-day hindcasts. This methodology allows the models to be evaluated against detailed observations such as at ARM sites, CloudSat passes, etc. for particular synoptic events.
- Member of the WGNE-WGCM Diagnostics and Metrics Panel. The panel acts as a focal point for the development of tools, metrics and diagnostic techniques for model evaluation.