update 4/2005 Johannes Karstensen

# What's going on at the moment?

This is an incomplete page about ongoing projects using OMP analysis. Please send you research to provide a base for discussion.

## Method

A rather important 'result' of the OMP analysis are the residuals. By default the OMP package shows only the mass conservation residuals as a kind of zero order approach for the quality of the results. However, there is a lot more information available from the residuals of all individual tracers (stored in the variable 'error' in OMP2).
Related to this is a project Matt Tomczak and his students are working on: How to improve the source water type definitions by looking on the spatial distribution of the residuals?
Assuming a region is dominated through a certain source water type and say the oxygen residuals are relative high within this region one may change the oxygen source water values in such a way that the oxygen residuals are lower. Hence one could learn about the source water mass itself. Improvements of the residuals can be obtained in different way: A rather simple way would be to change the source water values by hand - but this can be an annoying procedure as the source water types are 'linked' together through the OMP matrix and solutions are found from a simultaneous fit to all parameter spaces. Hence changing one SWT (say oxygen) could change the residuals of other parameters as well. A more elegant way would be to solve the Hessian matrix. This turns out to be computationally very expensive. It also leads to multiple solutions and requires a good understanding of the situation to select the correct solution. An intermediate way, which is often also used on the input data to prove the stability of the OMP results, is to perform multiple simulations by adding controlled 'noise' to the source water matrix and analyzing its effect on the residuals of certain parameters. Matthias Tomczak and his students are currently exploring these possibilities.
Aneurin Henry-Edwards has developed a derivative of the OMP which he calls the TROMP - time resolving OMP. The method allows to trace the water mass composition even in the case some source waters slowly change their characteristics. This is an important contribution as we all know source water to be (more or less) variabiable over time.

## Arctic Ocean

Matthew Alkire studies water mass mixing in the Beaufort Sea for his master thesis.

## Southern Ocean

• Vagner Duarte will analyse data from the Brazilian cruises in Bransfield Strait during 2003 and 2004 as part of his masters thesis. He is under the supervision of Mauricio Magalhaes Mata (FURG).
• Martin Lourey is studying nutrient depletion in the Southern Ocean. He wants to use extended OMP analysis to investigate cross-frontal mixing of water masses.

## Miscellaneous

• Ben McNeil uses OMP to calculate Redfield ratios for the world oceans from WOCE data.