Temporal scaling large

Biodiversity can also be examined at ecosystem level. This is of special concern for ecology, because, although attention to ecological problems is increasing, the choice of the appropriate scale to look at the events is problematic (Levin 1999). The spatial and temporal scales needed are often of such large size to make any reliable observation difficult. This is especially true for the sea, due to the lack of barriers which, therefore, makes any subdivision into ecosystems difficult. This is why any quantitative evaluation of biodiversity at ecosystem level is far fi-om being an easy task. 

In the real world, however, the objects we see in nature and the traditional geometric shapes do not bear much resemblance to one another. Mandelbrot [2] was the first to model this irregularity mathematically clouds are not spheres, mountains are not cones, coastlines are not circles, and bark is not smooth, nor does lightning travel in a straight line. Mandelbrot coined the word fractal for structures in space and processes in time that cannot be characterized by a single spatial or temporal scale. In fact, the fractal objects and processes in time have multiscale properties, i.e., they continue to exhibit detailed structure over a large range of scales. Consequently, the value of a property of a fractal object or process depends on the spatial or temporal characteristic scale measurement (ruler size) used. 

For complex fluids, the gap in the spatio-temporal scales between the smallest microstructures and the largest structures is much smaller than for simple fluids. Dzwinel and Yuen (2000b) have shown that by using moderate number of particles, we can simulate in two dimensions multiresolution structures ranging frommicellar arrays to the large colloidal agglomerates. 

The grain growth almost always observed during metamorphism results in a decrease in snow SSA. The rate of decrease greatly affects snow albedo and e-folding depth, considered over large spatial and temporal scales. The initial decrease is very fast, with a factor of 2 decrease in I to 2 days. Experimental and field studies have quantified the rate of decrease of snow SSA as a function of temperature and temperature gradient." In all cases, the best empirical fit of SSA decay plots was of the form ... 

Different behaviour at different spatio-temporal scales, ranging from small-scale concentration fluctuations in the turbulence inertial range to large-scale air movements in synoptic events affecting the entire troposphere. 

Simple dispersion behaviour - as described above - is scale-independent, so that, for example, Gaussian variances estimated from small-scale experiments in a wind tunnel are applicable in the prediction of the atmospheric dispersion of smoke from a large chimney. However, the specification of a dispersion environment (the planetary boundary layer (PBL), terrain, obstacles, closed spaces, etc.) and particular assets (i.e. people or property that we might wish to protect) introduces absolute spatial and temporal scales. 

Transport occurs at a variety of spatial and temporal scales. In atmospheric models, it is customary to distinguish between motions that can be resolved on the numerical grid of the model (often called large-scale ) and sub-grid scale processes (such as... 

When considering large spatial and temporal scales the transport of a concentration field by advection and molecular diffusion can be be approximately described by a diffusion equation with an effective diffusion coefficient. The main question then is to find an expression for the effective diffusivity as a function of the flow parameters and molecular diffusivity. A range of this type of problems are discussed in the review by Majda and Kramer (1999). Here we consider two simple examples of this problem in the case of steady two-dimensional flows with open and closed streamlines, respectively. 

Using this model, the dependence of particles (monomers) concentration decay, equal to (1-Q), on temporal scale of polymerization t can be constmcted. Such dependence in donble logarithmic coordinates for fonr values c is presented in Fig. 26. As one can see, the adduced dependences at large enough t can be described by the following relationship [51] ... 

---