Circulation and Intrusions Northeast of Taiwan: Chasing and Predicting Uncertainty in the Cold Dome

Author Posting. © The Oceanography Society, 2011. This article is posted here by permission of The Oceanography Society for personal use, not for redistribution. The definitive version was published in Oceanography 24 no. 4 (2011): 110–121, doi:10.5670/oceanog.2011.99.

For any model estimate, the portion of the variability that contains error contributes to the uncertainty, and variability unresolved by model estimates or measurements is pure uncertainty (Lermusiaux et al., 2006).
Real-time uncertainty predictions using ensemble methods with numerical models and error subspace concepts have been used in the coastal ocean since 1996 (Lermusiaux, 1999). Such computations have been successfully employed in many coastal regions (e.g., Lermusiaux, 2001;Auclair et al., 2003

oceaNogr aphic BackgrouND
Two key circulation features in the region of interest are (1) the cold dome, a cyclonic feature that appears primarily in summer and is associated with upwelling, and (2) Kuroshio intrusions onto the continental shelf (Lee and Chao, 2003;Liang et al., 2003). The cold dome is an intermittent but common feature, approximately 100 km in diameter, that is evident in sea surface temperature imagery during the summer, when surface thermal patterns shift frequently.

Doming of the isotherms and isopycnals
in the center of the dome can lead to cyclonic circulation and upwelling over the shelf (e.g., Chern and Wang, 1992;Liu et al., 1992;Tang et al., 1999Tang et al., , 2000. Although it is thought that flow up the canyons leads to cold dome formation, wind forcing may also play an important role (Chang et al., 2009). Further information on the processes in this area appears in Jan et al. (2011, in  northeast of Taiwan appears in Figure 1, and also in Jan et al. (2011, in this issue).

Glider observations were obtained
T at 50 m depth Aug. 13−17, 2009 The upwelling generated an east-west sea level difference, which in turn generated a northward geostrophic current.
Another feature present over the continental slope during the IOP was figure 3. (a) temporal variation of salinity averaged between σ θ 24 kg m -3 and 26 kg m -3 at six moorings deployed east and northeast of taiwan. (b) horizontal velocity averaged from the sea surface to either the bottom or 1,000 m depth from gliders in 0.1° latitude x 0.1° longitude grid points (red velocity arrows), typhoon morakot track and maximum wind speed (filled circles color-coded by maximum wind speed), and six mooring locations (colored stars) that are color-coded and numbered to correspond to mean salinity records in (a). (c) potential temperature and salinity plot before typhoon morakot (left), immediately after typhoon morakot (center), and several days after typhoon morakot (right). colored dots correspond with color-coding of moorings (b, stars) and mean salinity records (a). The kuroshio temperature/salinity (t/S) data shown (solid gray line) were collected by gliders, and the east china Sea slope water t/S data (solid orange line) by em-apex floats. (d) average vertical profile of salinity (black), potential temperature (red), and potential density (blue) for the kuroshio collected by gliders in okinawa trough (solid lines) and the east china Sea slope collected by em-apex floats (dashed lines). average depth associated with σ θ 24 kg m -3 and 26 kg m -3 for the kuroshio (gray) and the east china Sea Slope (orange) shown with colored circles. real-time forecaStiNg of oceaN circul atioN aND uNcertaiNt y During the IOP, real-time analyses and forecasts of regional ocean circulation, regional acoustic propagation, and uncertainty estimates were issued using many of the observational assets  In addition to the analyses and forecast fields, daily uncertainty forecasts were produced using the method of Error Subspace Statistical Estimation (ESSE; Lermusiaux and Robinson, 1999;Lermusiaux et al., 2002;Lermusiaux, 2006). For each uncertainty forecast, an ensemble of 50 model runs was used to produce ocean field statistics.
Suggestions for SeaSoar and acoustic sampling plans were provided daily based on these uncertainty fields.