| We carried out the numerical experiments by OGCM; MOM2.2 with usingreal topography database; ETOPO5 to reproduce the interannualvariability of the real Kuroshio transport at south of Japan; 1) PNline located in East China Sea, 2) ASUKA line located off CapeAshizuri. We adopted NCEP/NCAR re-analysis monthly mean wind stressdata as the driving force of OGCM in 1960-2000. Temporal variationpatterns between the OGCM′s transport and the observation′s oneroughly correspond until 1980s in PN line (correlation coefficient:about 0.6), though during 1990s are different in both PN and ASUKAlines (C.C.: less than 0.3). Since we considered that thisdisagreement might be caused from NCEP/NCAR data as one possiblereason, we re-drived OGCM by adopting another wind stress data; ECMWFand calculated the Kuroshio transport as same as the case of theNCEP/NCAR′s calculation. In this ECMWF calculation case, we adoptedthe data only in 1979-1994 because of a restriction of data existanceperiod. Temporal variation pattern of the Kuroshio transports during1980s between the case of using NCEP/NCAR and ECMWF data is verysimilar in both PN and ASUKA lines. On the other hand, variationpattern between NCEP/NCAR′s and ECMWF′s during 1990s is differentcompared with the 1980s in ASUKA line, that also denotes a littlecorrespondence with the observation′s transport. This consequencesuggests that one reason of the disagreement of temporal variationpattern of transports between observation′s and OGCM′s during 1990smight be caused from NCEP/NCAR data. Of course, there might beanother possible reasons that considered as follows; 1) variation ofintegrated Sverdrup Transport to the western boundary, which isgenerated by alteration of propagation velocity of first modebaroclinic Rossby Wave caused by the change of stratification degree,2) variation of the mode of baroclinic Rossby Wave at Izu Ridge, thatcause the variation of integrated Sverdrup Transport to the westernboundary, 3) non-linear effect such as the Kuroshio meandering. |
|
|