Page 208 - Fundamentals of Ocean Renewable Energy Generating Electricity From The Sea

Ocean Modelling for Resource Characterization Chapter | 8 197

the interisland channels are resolved using the high-resolution dataset. In more
advanced modelling systems, it is not necessary to clearly define a coastline, as
it will change by wetting and drying, and this moving boundary is computed
by the model. For those cases, a larger domain that includes bathymetry and
topography (e.g. 10 m above MSL) will be used as input.
Bathymetry Data
Although it is possible to generate a model grid using even the coarsest of
bathymetric data, it is generally desirable for the resolution of the bathymetry
data to match or exceed the model grid resolution. For example, there is little
point generating a model that has a grid resolution of 25 m if the available
bathymetry for the region only has a resolution of 1 km.
There are a wide range of sources of bathymetry data, which vary in
scale and resolution from multibeam surveys (resolution <10 m) to global and
regional datasets such as GEBCO (global 1/2 arc-min grid), NOAA bathymetry
and global relief, and EMODnet (European 1/8 arc-min grid). In general,
gridded bathmetry datasets tend to be suitable for shelf-scale or regional-scale
models, whereas high-resolution multibeam data may be required for refined,
localized, model studies. Finally, it is important to correct available bathymetry
data to the desired model datum, usually mean sea level (MSL). In general,
bathymetry data are available relative to the lowest astronomical tide, LAT 1
(see Section 3.9), and so accurate information on tidal range is required to
correct to MSL.
Boundary Data
Wave and tidal models require boundary conditions. The models may be nested
within a (coarser) outer model domain, which directly transfers information
from outer to inner nest (e.g. [2]). However, at some stage, and at some level
of nesting, a model will likely require boundary information from an external
source. For tidal models, many freely available ‘tidal atlases’ are suitable, such
as TPXO. Tidal atlases represent an assimilation between telemetry (satellite)
data, in situ data, and global numerical models. One such source, FES2014 [3],
is available at a (global) grid resolution of 1/16 × 1/16 degrees for 32 tidal
constituents, for both surface elevations and tidal currents, as both amplitude and
phase components (Fig. 8.4). For wave models, although there are various wave
products, which provide statistical wave properties (e.g. NOAA Wavewatch III
which provides Hs and Tz), it is generally preferable to seek products, or to run
simulations, which transfer 2D wave spectra from outer to inner nest (e.g. [4]).

Surface Forcing
Surface fields are required to force tidal models, which simulate nonastronomi-
cal processes (e.g. wind-driven currents or surge), and wind fields are generally

1. LAT is traditionally used on navigational charts, because it represents the shallowest possible
water depth.