Estimating the change in global sea levels: Bayesian methods in inverse problems

strict warning: Only variables should be passed by reference in /var/www/html/drupal/sites/all/modules/content_profile/ on line 13.
Patrick Raanes
Oxford University, UK
Seminar Date: 
22. October 2012 - 11:15 - 11:45
Lecture room, Ground Floor, NERSC

Patrick Raanes is applying for the PhD position in data assimilation and oil reservoir. He will present his 1st year PhD and his summer internship:

"Estimating the change in global sea levels:
Bayesian methods in inverse problems"
Current estimates for global sea level rise over the 20th century are around 4cm, with some indication that this might have accelerated the last couple of decades. However, satellite data (Topex/Poseideon, Jason) only stretches 20 years back in time. The most recent estimates are based on the method of empirical orthogonal eigenfunction (EOF) reconstruction of the historic sea levels, where the EOFs are obtained from satellite observation data, and then fitted to older tide gauge data. Our approach is to use data assimilation methods such as the particle filter and the EnKF to get a more reliable estimate. The presentation will include mostly background theory.

"Reservoir souring in Eclipse:
Implementing tracer conversion"
Reservoir souring refers to the generation of H2S gas in oil reservoirs by bacteria injected with sea water. H2S gas is corrosive and highly toxic. We wish to simulate the process in Eclipse. Initial gap analysis indicates that Eclipse does not fully support this yet, as its chemical reaction facilities are not sufficiently broad, and somewhat inflexible. Therefore we implement a "proof of concept" tracer conversion scheme for the blackoil simulator that supports the simulation of reservoir souring. The presentation includes background, implementation and results.