BISMG:TamsinE

8th January 2010
Spent some time downloading software onto the new laptop, organising all my pdfs of articles, organising all my paper copies of articles, unpacking more stuff in the office. Currently more organised than I have ever been, due in part to two excellent pieces of software: 'Papers' for, er, papers, and 'Evernote' as an electronic logbook (I use a hardback logbook too).

Started reading ensemble Kalman filter papers, beginning with the method section of Hargreaves & Annan (2002), "Assimilation of paleo-data in a simple Earth system model". They use Monte Carlo Markov Chain (essentially correlated random walk) to estimate pdfs of paramaters for Saltzmann and Marsch model of glacial-interglacial cycles. Found the notation confusing at first - they appear to use the index 'i' to mean two different things (number of observations and current set of parameter values) and similarly with prime (model output rather than observations, and trial set of parameter values).

Will work my way through other Annan and Hargreaves papers from Tony in date order. Jonty has doubts about the use of a Kalman Filter if the parameters are not stochastically varying through time (as in his work with Michel) - need to ask him again about this to make sure I understand.

November - December 2009
I'm officially working for 1 day a week until the end of January, then 2 days a week in Febuary, then full-time from March. Due to a slight overload I'm nudging some of my November days into December and January.

I intend to spend the last few days before Christmas catching up on some ice2sea-related work or study. This will bring me up to about 6 days total, and I'll carry 2 days over into Jan/Feb.

I've been to these meetings and courses:


 * 20/11/09 ice sheet course
 * 13/11/09 JCRP @ UKMO
 * 11/11/09 BISMG
 * 27-28/10/09 ice2sea WP5 @ BAS
 * 22/9/09 Tony and Jonty