Accurate descriptions and forecasts of climate variabilities are of great importance to a wide cross section of society. Toward the better representation of the climate variabilities such as Pacific Decadal Oscillation (PDO), dynamical state estimate has been done by using a 4D-VAR data assimilation system. The reproduced climate variabilities, including PDO, reflect most of the familiar gross feature of past studies (Figure 1). Further, using the obtained dataset, water mass transport was estimated at 24oN in the Pacific Ocean (Figure 2). The obtained result shows clear dependence of the time scale of the variability on the depth of the corresponding isopycnal layer.

Figure 1. (Upper panel) Time series of the equatorial SST in the Pacific Ocean; assimilation (green curve) and observation (back). (Lower panel) Time series of PDO index (from http://jisao.washington.edu/pdo/PDO.latest ).		Figure 2. Time series of estimated mass transport at 24°N in the Pacific Ocean during 1987-2004; Thermocline layer (blue curve), NPIW layer (green), and AAIW layer (pale blue).