Using OFES, first calculation and sensitivity to method, final distance, time of integration: little sensitivityOne year of stable and unstable manifolds in the northeastern equatorial Pacific in OFES QUICKSCAT run: zonal band of high FSLE at the latitude of Hawaii, annual cycle with a spring peakSeven years of stable and unstable manifolds in the northeastern equatorial Pacific in OFES QUICKSCAT run: confirmation of an annual cycle of FSLE with a peak in spring-summer. Is this related to the summer blooms observed around Hawaii?Sensitivity of FSLE events to SSH products: The FSLE are sensitive in the details to the SSH product used. However, its large-scale pattern is not too sensitive.

Comparison of unstable manifold and deep Chl maximum at station ALOHA: No obvious correlation between FSLE and deep CHl maximum. However, this needs to be redone: At this time, I was using the SSH anomaly not the absolute one. Some differences may arise.FSLE and upwelling events along float trajectory - Part II: There is a mistake somewhere. I am in the process of redoing the calculation.

FSLE and vertical velocity in Paulo’s model: good correlation at the large-scale between vertical velocity andf backward-in-time FSLE. Weak or no correlation at the submesoscale which makes it difficult the analysis of a single location.FSLE with and without submesoscale flow: This shows that the mesoscale flow alone is responsible for the submesoscale structure as calculated by FSLE. In other words, the present T/P SSH product is sufficient to localize the submesoscale structure.

- the large-scale pattern of FSLE is not too sensitive to either the method, the parameters or the SSH product used. However,it is sensitive at the submesoscale so care is needed when one focuses on a particular position.
- region of enhanced FSLE around Hawaii together with the spring-summer peak could explain the summer blooms.
- so far, not successful in showing a good correlation between backward-in-time FSLE and either deep Chl maximum at station ALOHA and nitrate upwelling events observed from the float.
- this failure is in part due: 1) other processes, in particular the mesoscale that does play a direct role in the ecosystem around Hawaii and the biological processes that make the nitrate and the deep Chl maximum not pure passive tracers, 2) weak correlation (even in numerical model) between vertical velocity and nitrate (see Riviere and Pondaven 2006; their FIg. 6), 3) the difficulty in focusing on one single location.
- the use of a high resolution numerical model could help to analyse the observations by giving robust methods to link FSLE, vertical velocity, nitrate concentration and deep Chl maximum.

- relationship between FSLE and DMS gas exchanges: work with Barry Huebert and Ming
- relationship between FSLE and tuna/turtle migration: work with Jeffrey Polovina and Reka Domokos
- remove the effect of submesocale using Martin and Pondaven (2003)
- use of the surface Quasi-geostrophy theory as an alternate estimate of vertical velocity for the observations at station ALOHA and from float
- use of a high-resolution numerical simulation with biological component forced from the east by a realistic mesoscale flow: 1) effect of downscaling, 2) correlation between nitrate concentration, deep Chl maximum, FSLE, vertical velocity, etc.
- analysis of the 1/10 and 1/30-° OFES run: differences, relation between FSLE, deep Chl maximum, FSLE, vertical velocity, etc around Hawaii and in the tropics.