| Abstract: | The development of hydrodynamic numerical models for environmental studies depends on good benchmarks to calibrate and validate the physics and numerical codes. Laboratory models of non-linear and coupled physics in topography for which no analytical solutions are available can provide such valuable benchmarks. Although field data are necessary for a final validation, they are often of less value for developing numerical models, since a truly synoptic coverage of a scenario is seldom found, knowledge of the forcing conditions is imperfect and average conditions of a non-linear system are seldom obtained by applying average boundary conditions.The role of laboratory models and experiments for providing information on turbulence in complicated topography is indisputable. The high topographical resolution of these models reveals how narrow and filamentous many of the flow features can be, as often seen in satellite images. Such filaments enhance diffusion through a process known as shear dispersion. The filaments are also of concern for the interpretation of sparse field measurements and for computing the mesoscale (10–100 km) spreading characteristics. Time histories of dye clouds and clusters of particles in laboratory simulations of ocean currents, without wind, show much larger spreading than particle spreading due to strong winds. The results demonstrate that numerical models need high resolution and/or good parametrization of the spreading characteristics, which vary both in space and time, to achieve their goals. It is proposed that the differences between numerical and laboratory simulations of dispersion, with identical forcing, be parametrized as a size-dependent, or time-dependent random walk diffusion in the numerical code.The laboratory results amply show that spreading is greatly enhanced by shear dispersion, and that assessments of the consequences of accidental oil spills or releases of radionuclides, for example, must take this into account. Island communities in tidally active regions are particularly prone to the consequences of a rapid dispersion of contaminants. |
