Use of ATSR Data and In Situ Observations to Study Ocean Dynamics near the Kuroshio Boundary

Abstract

0.5° spatially averaged sea surface temperature (SST) from the along-track scanning radiometer (ATSR) onboard ERS-1 was compared with in situ data measured by CTD during three cruises in the East China Sea in 1993 and 1994. The mean difference between ATSR SST and CTD SST is -0.5 K and standard deviation is 0.37 K, while the co-registration between ATSR SST and in-situ SST measurement is within 12 hours temporally and 0.5° spatially. The monthly 0.5° averaged ATSR SST in the region 0-45N and 100E -145E from August 1991 to July 1995 was used to study the spatial and temporal variability of SST near the Kuroshio boundary. The four-year sequential SST data show dynamic surface patterns of the Kuroshio, as well as the current system near the Kuroshio boundary. A meander path of the Kuroshio occurred to the south of Honshu in 1992, 1993 and 1995, whereas a straight path appeared in 1994. Statistical analysis of SST along two typical transects (PN and TK, see figure 1), cross the path of the Kuroshio was carried out to examine temporal variability of SST. The results show that SST was relatively high in spring and summer 1994. SST observation along Peruvian coast indicates that 1994 marks the termination of 1991/1992 and 1992/1993 El Niño events. The relationships between SST anomaly in the area of the Kuroshio, the transition of the Kuroshio path and El Niño event are discussed.
Keywords: ATSR, SST, Kuroshio, El Niño

1. Introduction

The Kuroshio, western boundary current in the Pacific Ocean, plays important role both in the oceanic and atmospheric processes in the Pacific. The variability of the Kuroshio induces changes in atmospheric processes through air-sea interaction and hence influences the regional as well as global climate. The bimodality of the Kuroshio path to the south of Japan is one of the most remarkable interannual variations in the North Pacific [Akitomo, et al., 1996]. That is, there are two possible stable paths of the Kuroshio to the south of Houshu, Japan, a meander and a straight path. The mechanism of the bimodality of the Kuroshio path is far from clear. Previous studies have shown that change in the volume transport of the Kuroshio is associated with the change in the Kuroshio path. Yutaka [1992] reported that large volume transport causes a straight path and small volume transport causes a meander path. While Ichikawa and Beardsley [1993] believed that volume transport in the East China Sea is large when the Kuroshio takes a meandering path to the south of Japan. Akitomo et al. [1996] found that the variation of the wind stress curl anomaly associated with the ENSO event can play an essential role in the interannual variability of the Kuroshio, the negative anomalies of the wind curl due to the ENSO event result in the transition from a straight path to a meander path while the decaying process of a meander path is still unclear.

The along-track scanning radiometer (ATSR) onboard ERS-1 was launched in July 1991, designed to provide very accurate sea surface temperature (SST). ATSR has four channels at 10.8 m, 12 m, 3.7 m and 1.6 m, with a spatial resolution of 1 km at nadir. It incorporates several novel features which enable it to meet performance objective, including a dual angle view of the earth's surface, two stable and high-precision on-board blackbody calibration targets, a Stirling cycle cooler to maintain the detectors at a temperature of around 90 K, and 12-bit resolution of the raw data transmission [Delderfield et al., 1985]. The early validation results from the comparison between 0.5° spatially averaged global ATSR SST and drifting buoy data show that SST derived from dual-view three-channel algorithm has a bias of -0.03 K and a RMS deviation of 0.36 K against buoy data, and that SST from dual-view two-channel has a bias of -0.36 K and a RMS of 0.42 K [Multow et al., 1994].

In this paper, preliminary validation of ATSR SST against in situ data in the East China Sea is carried out. ATSR SST is used to study spatial and temporal variability of SST near the Kuroshio boundary and to investigate the dynamics of the Kuroshio. The mechanism of the transition of the Kuroshio path is discussed.

2. DATA

The 0.5° spatially averaged ATSR SST is retrieved as skin SST. Mean brightness temperature in a ten-arcminute cell is used to retrieve a ten-arcminute SST with the coefficients derived from a radiative transfer model. Then all the available (up to nine) ten-arcminute SSTs are averaged [Závody et al., 1994]. ATSR SST with associated time and confidence information was extracted from global ATSR daily spatially averaged SST products and compared with CTD data. Monthly and spatially averaged ATSR SST during the period from August 1991 to July 1995 was used to study the temporal and spatial variability of SST near the Kuroshio boundary.

Three cruises for the program on Margin Fluxes in the East China Sea (MFLECS) have been carried out in the East China Sea during the period from 15 October to 28 October 1993, 8 April to 30 April 1994 and 18 October to 13 November 1994. CTD raw data measured during the cruises were converted to physical values and averaged for each 1 m interval using a software package SEASOFT version 4.26 on real time [Cui et al., 1995]. Temperatures at the minimum pressure were chosen to compare with ATSR SST. In the match-up data set, most of CTD temperatures were at pressure 0db, two cases on pressure 5db. Hence, temperature from CTD that was used to compare with ATSR SST is regarded as bulk SST.

Wind stress data from the United Kingdom Meteorological Office (UKMO) cover the area 123°E-69°W and 30°S-30°N on 2°2°grids, for a period from August 1991 to April 1995.

3. RESULTS AND DISCUSSION

3.1. Comparison of ATSR data with in situ data

ATSR SST data, within less than 24 hours and less than 0.5 of corresponding CTD measurements, were extracted. CTD SST data in the same 0.50.5 cell were averaged. The results are shown in table 1. The mean difference between ATSR SST and CTD SST is -0.53 K and the standard deviation is 0.51 K. While match-up data with temporal difference more than 12 hours were rejected, the mean difference is -0.50 K and the standard deviation is 0.37 K. Considering possible error sources, firstly, the temporal and spatial difference between the ATSR and CTD measurements attribute to the error. Secondly, CTD SST is bulk temperature while ATSR SST is skin temperature. The skin effect can lead to the bulk SST of several tenths of a degree lower than the skin SST.

Table 1. SST difference between ATSR and CTD.

3.2. Study of ocean dynamics near the Kuroshio boundary

The monthly 0.5° averaged ATSR SST data in the region 0°- 45°N and 100°-145°E from August 1991 to July 1995 were examined to study the spatial and temporal variability of SST near the Kuroshio boundary. The four-year sequential SST is shown in plate 1. The influence of the Kuroshio on SST distribution is evident. Plate 1 shows the path of the Kuroshio as well as mesoscale oceanic phenomena that occurred near the Kuroshio boundary. The SST spatial distribution is more homogeneous in summer than that in other seasons due to strong solar radiation. Even in summer, the Kuroshio path can be seen from ATSR SST (August 1993). Filaments and eddies near the Kuroshio boundary were detected by ATSR SST. The current system near the Kuroshio boundary in winter is clear in Plate 1. The warm tongues reflect Taiwan Warm Current, Huanghai Warm Current and Tsushima Warm Current in China Seas. The origins of the above mentioned warm currents have certain relationship with the Kuroshio. A cold tongue reflects runoff from Changjiang River, the largest river in China. The information demonstrates the ability of ATSR monthly averaged gridded SST to detect mesoscale phenomena.

Plate 1. Monthly averaged SST from ATSR in the region 0º-45ºN and 100ºE-145ºE for the period August 1991 to July 1995

Figure 1. Study area. PN and TK are transects cross the Kuroshio path

An intriguing phenomenon revealed in plate 1 is the interannual variability of the Kuroshio path. The surface path is clear in each May during the four-year period. To investigate whether SST is representative of the Kuroshio path, temperature profiles from CTD were examined. CTD data along PN (see figure 1), measured at 16 stations during two cruises, one from 2156 GMT 15 April to 0353 GMT 17 April 1994, and the other from 0240 GMT 31 October to 1623 GMT 1 November 1994, are shown in figure 2. Data at station 8 during the latter cruise were absent. Temperature at 0db, 10db, 20db, 30db and 50db increased significantly from station 8 to station 9 and had similar variation trend in April, while the horizontal temperature gradient was not significant at the above depths in the latter cruise. CTD data indicate that SST can be a good indicator for the path of the Kuroshio in Spring and not a good indicator for that in Autumn. Thus ATSR SST can be representative of the upper layer of the Kuroshio in May. Meander and straight paths concurred to the south of Honshu, Japan in May 1992, which the former is more strong. Meander path persisted in May 1993. While SST in May 1994 exhibits a straight path to the south of Honshu. Meander path appeared again in May 1995 (see plate 1).

Figure 2. CTD temperature profiles along PN (a) from 2156 GMT 15 April to 0353 GMT 17 April 1994; (b) from 0240 GMT 31 October to 1623 GMT 1 November 1994.

Figure 3. Temporal variability of SST (a) averaged along PN; (b) averaged along TK; (c) averaged in the region 10N-20N and 130E- 140E; (d) averaged along Peruvian coast.

SSTs along two transects cross the Kuroshio path, PN and TK (see figure 1), were averaged. Figure 3(a) and figure 3(b) show that averaged SSTs at PN and TK were much higher in spring and summer 1994 by more than half a degree, compared with the other three years. SST averaged in the region 0°-20°N and 130°E-140°E, the region of the Kuroshio origin, is shown in figure 3(c). Warmer SST appeared in November 1993 and persisted till June 1994. The results indicate that warm SST occurred near the boundary of the Kuroshio in 1994. It has been reported that El Niño occurred in 1991/1992 and 1992/1993 [Bell et al., 1994; Kessler et al., 1995]. SSTs of five sites along Peruvian coast were averaged. The centers of each site locate at 4.75°S and 81.25°W, 7.75°S and 79.25°W, 9.25°S and 78.75°W, 10.75°S and 77.75°W, 11.75°S and 77.25°W. Figure 3(d) shows that warm northern winter appeared near Peruvian coast in 1991/1992 and 1992/1993 and cold northern winter appeared in 1993/1994, which is consistent with the reports.

Wind stress curl during the period from August 1991 to April 1995 was investigated. Emphasis was laid on the interannual variability. Annual changes in each month were examined. The result reveals that wind stress curl in December 1993 was much higher than those in other years.Plate2 shows wind stress curl difference between each December in the region of the Kuroshio origin, to the east of Philippine.

Plate 2. Wind stress curl difference between (a) Dec 1992 and Dec. 1991; (b) Dec. 1993 and Dec. 1992; (c) Dec. 1994 and Dec. 1993. Value is scaled between ±2.5×10^-7Nm^-3

The results show coincidence of significant warm SST near the Kuroshio boundary, transition from a meander path to a straight path to the south of Japan and high positive wind stress curl in the region of the Kuroshio origin when it returned to a normal year from an El Niño year. It indicates that there is relationship between El Niño event and interannual variability near the Kuroshio boundary. A possible mechanism for the phenomena in 1993/1994 is presented here. After termination of El Niño event, warm water exits in the western tropical Pacific. Wind stress strengthens the Kuroshio and warm water propagates from tropical area to mid-latitude area. It seems that large transport caused a straight path in 1994.

4. Conclusions

The mean difference and standard deviation between ATSR SST and CTD SST are -0.5 K and 0.37 K respectively, within 12 hours temporal limit and 0.5° spatial limit. Four-year sequential ATSR SSTs indicate some relationship between the transition of the Kuroshio path, SST anomaly near the Kuroshio boundary and El Niño during the period from August 1991 to July 1995. Straight path to the south of Japan appeared in 1994 when it transited from an El Niño year to a normal year. SST was higher near the Kuroshio boundary in spring and summer 1994, compared with other three years. The results suggest that SST anomaly near the Kuroshio boundary is associated with El Niño event. Analysis of wind data suggests that El Niño event may affect the Kuroshio path via wind stress. Further investigations using long-term satellite data, in situ data combined with numerical model are required to better understand ocean dynamics near the Kuroshio boundary.

Acknowledgments. The authors would like to thank ATSR team at Rutherford Appleton Laboratory for providing ATSR data and the United Kingdom Meteorological Office for wind stress data. MFLECS project is funded by National Science Foundation Commission of China. Lei Guan is supported by an European Space Agency Fellowship at the University of Leicester.

References

Akitomo, K., M. Ooi, T. Awaji, and K. Kutsuwada, 1996:

Interannual variability of the Kuroshio transport in response to the wind stress field over the North Pacific: It's relation to the path variation south of Japan. J. of Geophys. Res., 101, pp. 14057-14071.

Bell, G. D., and A. N. Basist, 1994:

Seasonal climate summary. The gloabal climate of Dec. 1992-Feb. 1993. Part I: Warm ENSO conditions continue in the tropical Pacific; California drought abates. J. of Climate, 7, pp. 1581-1605

Cui, M., and H. Bai, 1996:

CTD data acquisition and processing. Report on the program on Margin Flux in the East China Sea, Institue of Oceanology, Sciences Academy of china, p1.

Delderfield, J., and D. T. Llewellyn-Jones, R. Bernard, Y. de Javel, E. J. Williamson, I. Mason, D. R. Pick, and I. J. Barton, 1985:

The along track scanning radiometer (ATSR) for ERS-1. Proc. SPIE, Int. Soc. Opt. Eng., 598, pp. 114-120.

Hurlburt, H. E., A. J. Wallcraff, W. J. Schmity Jr., P. J. Hogan, and E. J. Metzger, 1996:

Dynamics of the Kuroshio/Oyashio current system using eddy-resolving models of the North Pacific Ocean. J. of Geophys. Res., 101, pp. 941-976.

Ichikawa, H., and R. C. Beardsley, 1993:

Temporal and spatial variability of volume transport of the Kuroshio in the east China Sea. Deep Sea Res., 40, pp. 583-605.

Kessler W. S., and M. J. McPhaden, 1995:

Oceanic equatorial waves and 1991-93 El Niño. J. of Climate, 8, pp. 1757-1774.

Mutlow, C. T., A. M. Závody, I. J. Barton and D. T. Llewellyn-Jones, 1994:

Sea surface measurements by the along-track scanning radiometer on the ERS1 satellite: Early results, J. of Geophys. Res., 99, pp. 22575-22588.

Yutaka, N, 1992:

Review of recent studies on the Kuroshio, Proceedings of China-Japan joint synposium of the cooperative research on the Kuroshio. China Ocean Press, Beijing, pp. 8-17.

Závody, A. M., M. R. Gorman, D. J. Lee, P. Bailey, D. Eccles, C.T. Mutlow, and D. T. Llewellyn-Jones, 1994:

The ATSR data processing scheme developed for the EODC. Int. J. Remote Sens., 15, pp. 827-843.

Keywords: ESA European Space Agency - Agence spatiale europeenne, observation de la terre, earth observation, satellite remote sensing, teledetection, geophysique, altimetrie, radar, chimique atmospherique, geophysics, altimetry, radar, atmospheric chemistry