ABSTRACT
Experiments have always played an important role in advancing scientific understanding. While many aspects of environmental flows are “observable,” the researcher lacks the ability to separate different elements and processes, which is often necessary to gain the key understanding, test theoretical and numerical models, or make predictions based on “what if” scenarios. Often we have a good idea of the equations describing the physics, but these are too complex to solve analytically and we are forced to utilize simplified systems of “model” equations. Numerical simulations entail approximate solutions to approximate sets of equations: we cannot model directly all the scales of turbulence in any real flow, or even begin to represent accurately the geometry of a porous medium. These limitations do not mean that theoretical and numerical activities have no value, but rather that at least some of their predictions and conclusions need to be tested against observations and against the more controlled conditions found in the laboratory.
