Therefore, the dynamics of the density current are considered to be similar to the theoretical results by Kubokawa and Hanawa (1984b). The propagation speed of the leading edge obtained by their model is estimated as approximately C = (1.45±0.02)(g'h)1/2. Then its speed is estimated as C = 0.7 ms-1 applying (g'h)1/2 = 0.45 ms-1 obtained from the observations. The speed is easily confirmed as nearly equal to the propagation speed, 0.6-0.7 ms-1, of the warm water estimated by the time lag and distance between the observation sites. Kubokawa and Hanawa (1984b) also indicated the existence of undulations after passing of the warm front. The time series of current records at both sites (Fig. 3) suggest the fluctuations after the strong current and passing of the warm front. Then, the observed strong current with the thermal front, i.e., warm water intrusion, is consistent with the coastal density current.
4.2. Variabilty of the Kuroshio
The current in the upper layer of Sagarni Bay is strongly influenced by the fluctuations of the Kuroshio Path (Taira and Teramoto, 1986; Kawabe and Yoneno, 1987; Iwata and Matsuyama, 1989; Teramoto, 1996). All data, including the current and temperature data at moored and tower stations, CTD data and NOAA IR-images, indicate the Kyucho is induced by the sudden approach of the Kuroshio to Sagami Bay. So, we investigate whether there were the remarked variations of the Kuroshio path before and after the Kyucho. Their variations can be found from prompt reports of hydrographic observations in Fig. 8 (Japan Maritime Safety Agency). The horizontal temperature distributions at 200 m depth were drawn at every half month from early December 1993 to late January 1994. The Kuroshio axis was reported to agree well with the 15℃ isotherm at 200 m depth south of Japan statistically (Uda, 1964; Kawai, 1969; Taft, 1972). In addition, as seen in Fig. 8, the 15℃ isotherm is nearly the maximum of the temperature gradient across the Kuroshio. The Kuroshio path estimated from the 15℃ isotherm was variable before and after the Kyucho. The Kuroshio was in the nearshore non-large-meander path (nNLM, named by Kawabe, 1985) during December 1993, but changed to a very complicated path during the period of January 6-19, 1994. The warm water separated from the Kuroshio then moved westward along the Japan coast and formed a cold eddy. After this, the Kuroshio moved offshore together with the cold eddy, namely, to the offshore non-large-meander path(oNLM). The Kyucho occurred during the transition of the Kuroshio path from the nearshore to offshore non-large-meander paths.
Our observations show a sudden increase in water temperature and salinity along the coast of Sagami Bay indicating that warm and saline water originating from the Kuroshio intrudes into Sagami Bay. Consequently, this Kyucho is speculated to occur as the density current of the warm and saline water which was trapped along the coast during the variation of the Kuroshio path. Further understanding of the dynamics link between the Kuroshio meanders and the Kyucho events is an important problem which should be studied for prediction of the Kyucho.
