Lessons from high-resolution satellite SST's

Since satellite-derived sea surface temperatures (SSTs) are retrieved from measurements of emission from the ocean surface skin, their utility in characterizing the temporal and spatial variability of oceanographic features has been questioned. However, analyses derived from a 10-yr archive of 1-km resolution SST data-called the National Institute of Water and Atmospheric Research (NIWA) SST Archive (NSA)-are providing new insights into the complex physical oceanography of the southwest Pacific region.

For example, the annual mean (Fig. 1) reveals clear signatures of many of the important features in the region, including the mean locations of current systems and fronts. Further, these "skin-deep" SST measurements clearly contain information about the effects of vertical mixing and/or flow diversion caused by bathymetric features far below the surface (e.g., over the Kermadec Ridge). The amplitude and phase of the annual cycle (Fig. 2) also reveal additional detail about mixed layer stratification. Specifically, regions where the annual cycle has small amplitude and late phase seem associated with bathymetric features, while the absence of bathymetric features leads to more spatially uniform amplitudes. Exceptions are also evident, such as in the sub-Antarctic watermass south of the Chatham Rise, where the annual cycle displays an uncharacteristically large signal (and early phase), perhaps due to early summer stratification of the water column or advection of subtropical waters over the Chatham Rise. (A concomitant analysis of monthly mean chlorophyll-a data from the Sea-viewing Wide Field-of-View Sensor (Sea WiFS) mission, lends some support to the advection suggestion.) Regardless of the explanation, though, the spatially varying features evident in SST data are robust and appear in climatologies derived from both 5 and 10 years of data.

SST data also allow testable hypotheses to be posed. For example, the location of the Sub-Tropical Front (STF) east of New Zealand is still unknown, but the NSA suggests that it dips southward east of the Chatham Rise. Recent research voyages have found evidence to support this hypothesis.

A comparison of the NSA and NOAA Pathfinder SST reanalysis datasets indicates that the former is warmer (mean difference over the region is 0.216[degrees]C), but the location and magnitudes of all features, including those in the highly varying Antarctic Circumpolar Current (ACC), are identical. Given the difference in resolution of the Advanced Very High Resolution Radiometer (AVHRR) data sources utilized (4 km for Pathfinder, 1 km for NSA) and the completely different cloud-detection methods employed in the Pathfinder (decision tree) and NSA (Bayesian) data, this is an encouraging result.

Last, while some evidence of ocean warming has previously been reported in this region, a watermass-dependent analysis of SST anomalies over the region, and the period 1985 to the present, reveals only ENSO-related climate signals. There is no evidence of ocean warming.

[Author Affiliation]

-MICHAEL J. UDDSTROM (NATIONAL INSTITUTE OF WATER AND ATMOSPHERIC RESEARCH). "Ten Years of High-Resolution Sea Surface Temperatures-What Have We Learned?"