An observing system simulation experiment for the Western North Pacific region.

This study investigated the effectiveness of concentrated observations for ocean state estimation in a region remote from the observation site. I executed a twin observing system simulation experiment (OSSE) for the North Pacific region, using an ocean data synthesis system, to examine how the potential effectiveness is for a well-defined criterion, the representativeness of the subsurface salinity minimum corresponding to North Pacific Intermediate Water (NPIW). The results of the OSSE show that data synthesis confined to the region corresponding to the recent origin of the NPIW (35 °N-53 °N, 130 °E-170 °E) can affect the modeled extent of the NPIW in the central Pacific at 35 °N, 180°. The interannual variability of the NPIW is not well reproduced in terms of the standard deviation value (std), only by the data input in the origin region. The root mean square difference between the "true" and the synthesized field is twice larger than the std although there the representativeness of the scale of salinity minimum is improved by about one-third of the difference between the "true" and "first-guess" fields in a snapshot. These results imply that combinations of concentrated and other in situ observations should be required for the dynamic state estimation of the NPIW.