Summertime SST gradient distribution appear to beaffected rather by interannual variability, though it is irregular and does not match clear pattern of ENSO-scale variability in summertime SST anomalies.
Another important feature of structural pattern in ‘warm’ period is an appearance of secondary belt of increased SST gradient along 36N-37N, clearly separated from subpolar front by the low-gradient zone and associated with the Kuroshio extension. In zonally averaged data we analyze, this feature appears rather as a tendency, being obvious only in 1991-93. However, analysis of individual weekly SST gradient maps proved that Kuroshio extension appear in wintertime as a separate front throughout the most of 'warm' phase. Spatial distribution of 3-month mean SST gradient for 1984-85 and 1992-93 winter seasons (Fig.3) illustrates the differences in frontal structure between ‘cool’and ‘warm’periods (the maps have been calculated from interpolated SST products with linear trend in SST gradient removed).
CONCLUSIONS
In general, information on decadal-scale variability of SST anomalies in the mid-latitude North Pacific obtained from satellite data was consistent with previous studies based on ship measurements, both in spatio-temporal distribution and in magnitude. Satellite data proved to be useful for documenting the SST gradient variability in OFZ associated with decadal climatic shift. However, this data should be used with caution in climatological studies since the existing long-term trend in the number of missed pixels filled with interpolation, create artificial trend in SST gradient magnitude. One of the reasonable approaches to avoid this problem is decreasing the original resolution with spatio-temporal averaging.
Both subtropical and subpolar OFZ reveal changes in spatial structure and SST gradient anomalies pattern, associated with the decadal climatic shift of late 1980s. The decadal-scale variability in OFZ was more pronounced in wintertime (in fact, from late autumn until early spring), which agrees with previous findings (Nakamura et al. 1997). The reason for this may be the fact that in wintertime, due to the strong mixing, SST better represents the fundamental deep-water structure of frontal zones influenced by long-scale variability. In summertime, shallow seasonal mixed layer affected rather by the atmospheric interannual/ENSO variability tends to mask longer-scale signal.
