Under the currently funded JHT and HFIP projects, the URI group is in the process of completing the implementation of explicit wave coupling in the GFDL and HWRF systems, and its near-real time testing will begin in 2014. This model development effort involves 1) improving physical parameterizations of the air-sea heat and momentum fluxes at and near the sea surface with fully coupled wind-wave-current interaction and sea spray effects and 2) forging a comprehensive, scientifically integrated atmosphere-wave-ocean framework that couples individual model components.
At the heart of the coupled framework developed under the NOPP funding is a computationally efficient, unified Air-Sea Interface Module (ASIM) that establishes a consistent, physically based representation of the air-sea interface. In the ASIM, the bottom boundary condition of the atmospheric model incorporates sea-state dependent air-sea fluxes of momentum, heat, and humidity, and it includes the effect of sea-spray. The wave model is forced by the sea-state dependent wind stress and includes the ocean surface current effect. The ocean model is forced by the sea-state dependent wind stress and includes the ocean surface wave effects (i.e. Coriolis-Stokes effect, wave growth/decay effect, and Langmuir turbulence effect).
A key requirement for the ASIM is that it supports both technical and scientific interoperability over a range of models, parameterizations, and data resources. As an example, we illustrate here two approaches to the sea-state dependent drag coefficient parameterizations implemented into ASIM. The first approach (URI) is an extension of the approach developed by Moon et al. (2004). It has been updated by improving the spectral tail parameterization (the unresolved high frequency part of the wave spectrum), based on recent observational and theoretical findings (Reichl et al. 2012). The second approach (Univ. of Miami) is developed by Donelan et al. (2012). The two approaches are different in the following two areas: (a) the growth rate is parameterized based on the wind stress in the URI approach but based on the wind speed in the UM approach and (b) inside the wave boundary layer, the mean wind profile is modified (i.e. it is not logarithmic and may rotate) in the URI approach but, it is logarithmic and does not rotate in the UM approach. Figure 1 compares different drag coefficient parameterizations implemented in ASIM with fetch dependent seas using the empirical wave spectrum of Elfouhaily et al. (1997).
Figure 2 illustrates the URI and UM approaches with surface wave spectra simulated by WAVEWATCH under tropical cyclone conditions. The drag coefficient values are quite similar and the sea state dependence is relatively weak in both approaches. The UM result shows a slight increase of the drag coefficient in the left rear quadrant, where the sea is less developed. A notable difference between the two approaches is the misalignment angle between the 10-meter wind speed and the wind stress. While the URI approach always predicts small misalignment less than 2 degrees, the misalignment angle may exceed 4 degrees by the UM approach, particularly inside the radius of maximum wind.