Local and global estimations of recoverable reserves are often insufficient at the planning stage of a new mine or a new section of an operating mine. For the mining engineer, as well as the metallurgist and chemist, it is often essential to be able to predict the variations of the characteristics of the recoverable reserves at various stages in the operation, e.g., after mining, hauling, stockpiling, etc.
For instance, the choice of mining and haulage methods depends in part on the spatial dispersions of the various ore characteristics. Conversely, the potential recovery rate of the in situ resources will depend, in part, on the mining technology. The choice of mining equipment or of a method of removal of the mined products in a subhorizontal sedimentary deposit may depend on the daily variations in overburden and mineralized thickness. The blending process and the flexibility of the plant will depend on the dispersion of the grades received at all scales (daily, monthly, yearly).
If the in situ reality were perfectly known, the required dispersions, and, thus, the most suitable working methods, could be determined by applying various simulated processes to this reality. Unfortunately, the perfect knowledge of this in situ reality is not available at the planning stages of an operation. The information available at this stage is usually very fragmentary, and limited to the grades of a few samples. The estimations deduced from this information-even through kriging-are far too imprecise for the exact calculations of dispersions that are required.
As it is impossible to estimate the in situ reality correctly in sufficient detail, one simple idea is to simulate it on the basis of a model. In a way, a real deposit and simulations of this deposit are different variants of the same mineralized phenomenon, as characterized by a given model.
Consider, for example, the true grade variable
at each
point 
of a
deposit. The geostatistical approach consists in interpreting the spatial
distribution of the grade
as a particular
realization of a random
function 
. This random function is characterized
by its first two moments and its
distribution function, which are estimated from the experimental data. This
model is, thus, suitable for the practical problem of determining various
dispersions of the grades in the deposit, since the dispersion variances of
can be expressed as a function of the second-order moment
only-covariance or variogram. A simulation thus consists of
drawing another realization
of this random function
. The two realizations, real
and simulated, differ from each other at given locations but come from the
same random function , the first two moments and
the univariate distribution function of which are fixed.
As far as the dispersion of the simulated variable is concerned, there is
no difference between the simulated deposit
and the real deposit
.
The simulated deposit has the advantage of being known at all
points and not only at the experimental location points 
. This simulated
deposit is also called a ``statistical model'' of the real deposit.