Thus, energy produced by the gravitational attraction from the Moon is dissipated mainly by turbulent mixing, which occurs on the one hand via the bottom drag[2][3] and the eddy viscosity[2] on shelves, and on the other hand via the form drag[6] and the internal-tide breaking in deeper ocean[2][3]. The dissipation due to the form drag is termed (barotropic-to-) baroclinic conversion[7] and can be detected using data assimilative models by computing dynamical error work[8] or dynamical residual power[2] appearing in the deeper ocean > 1000 m (Fig.4). Highly accurate satellite altimetry thus allows us to look into the inside of the ocean via determining tidal currents and surface tides.

Exciting story: In the last decade, one has come to regard tidally induced mixing as an important driving force for the abyssal circulation[9] and hence to affect climate[10]. The investigation of this mixing requires good knowledge of tidal dissipation. Especially the baroclinic conversion rate[2] describing energy transfer from surface to internal tides, is hard to estimate, since the parameterization of the conversion is still not sufficiently established. This rate has been assumed to be about 1 TW, which is obtained from simple advective-diffusive balance[9], and from data assimilative models3). The estimation from the latter method is sensitive to the choice of bottom drag and eddy viscosity[2]. We obtained a new conversion rate that is by ~40% larger than previous investigatios[3], where dissipation parameters were estimated a priori within a reasonable range by sensitivity analysis. We found that MAR is the most effective energy conversion site due to its huge area (Fig. 4). The strong energy fluxes along MAR (Fig.3) may support our suggestion. The Weddell Sea and the NWES are coming to appear as significant conversion sites, which other models[3][4] did not detect so far as such.

Pathways of internal tides and their dissipation sites are nearly completely unknown. We suggest that adequately evaluating the dynamical residual power[2] distribution may help to determine a realistic vertical diffusivity.

**Epilogue**: ‘When asked which is more important, the Moon or the Sun? . The Moon, of course, because the Sun shines only in the daytime when it is bright anywhere[10]…‘.

PS: Tides and Current data of 9 constituents[11][12] are given free to download (icdc.zmaw.de/hamtide.html).

*[1] Bosch, B., R. Savcenko, F. Flechtner, C. Dahle, T. Mayer-Gürr, D. Stammer, E. Taguchi, and K-H. Ilk (2009), Residual ocean tide signals from satellite altimetry, GRACE gravity fields, and hydrodynamic modeling, GJI, 178(3), 1185-1192. doi :10.1111/j.1365-246X.2009.04281.x :
*

*
*[2] Taguchi, E., D. Stammer and W. Zahel (2011), Inferring deep ocean tidal energy dissipation from the global high-resolution data-assimilative HAMTIDE model (submitted to J. Geophys. Res.).

[11] Taguchi,E. and D. Stammer (2011), Ocean tides obtained by the data assimilative HAMTIDE model: 1. Harmonic constants of 9 major tides. ICDC.

[12] Taguchi, E. and D. Stammer (2011), Ocean tides obtained by the data assimilative HAMTIDE model: 2. Harmonic constants of 9 major tidal currents. ICDC.

*[3] Egbert & Ray (2001), JGR, 106: [4] Simmons et al. (2004), Dee-Sea R., 51: [5] Ray & Mitchum (1996), GRL, 23: [6] Zaron & Egbert (2006), JPO, 36: [7] Simmons et al. (2004), Dee-Sea R., 51: [8] Egbert (1997), PO, 40: [9] Wunsch & Ferrari (2004), A.R.F.M., 36: [10] S. R. Jayne (2009), JPO,39: [10] Munk & Wunsch (1997), Oceanogr. 10.*