next up previous contents
Next: The Variogram Up: Regionalized Variables Previous: Example 3.1: Deposit of   Contents


Computation of a Simple Variogram

Let us consider the simplest case, that values of a regionalized variable only in one direction in space are at hand. For illustration we compute by hand an empirical variogram from a few values only. Let us denote the support points which have distances $d$ by $x_{i}$, and the corresponding values of the regionalized variable by $z(x_{i})$; then the data may be as seen in Figure 3.2.

Figure 3.2: Regionalized Variable.
/begin{figure}
% latex2html id marker 4516
/setlength{/unitlength}{1mm}/begin{pi...
...x(0,0)[rt]{$26$}}
/put(132,5){/makebox(0,0)[rt]{$21$}}
/end{picture}/end{figure}

The data are taken from samples of grades of copper from Illustration (v) of Subsection 2.2.7 (Table 2.7), the first 15 values. The computation of the variogram is done by averaging over all available pairs of data $n(h)$, namely

/begin{displaymath}2 /hat{/gamma}(h)=/frac{1}{n(h)} /sum_{i=1}^{n(h)}
(z(x_{i})-z(x_{i}+h))^{2}/end{displaymath}

with $h=1 /times d, 2 /times d, 3 /times d, /ldots$. A ``manual'' computing scheme could look like in Table 3.1. For each selected $h$ a row of differences of the values is presented, and underneath the corresponding values squared. The sums, the number of used pairs $n(h)$ and the computed values $2 /gamma(h)$ are at the right margin. The resulting semi-variogram is shown in Figure 3.3.



Table 3.1: Calculation of a Simple Variogram.
  $x_1$   $x_2$   $x_3$   $x_4$   $x_5$   $x_6$   $x_7$   $x_8$   $x_9$   $x_{10}$   $x_{11}$   $x_{12}$   $x_{13}$   $x_{14}$   $x_{15}$ $/sum$ $n(h)$ $2
/hat{/gamma} (h)$
$z(x_{i})$   28   25   27   25   21   25   11   16   11   17   27   26   25   26   21
$h=50$                              
$z(x_{i})-z(x_{i+1})$   3   -2   2   4   -4   14   -5   5   -6   -10   1   1   -1   5  
$(z(x_{i})-z(x_{i+1}))^{2}$   9   4   4   16   16   196   25   25   36   100   1   1   1   25   459 14 32.8
100                              
$z(x_{i})-z(x_{i+2})$   1   0   6   0   10   9   0   -1   -16   -9   2   0   4    
$(z(x_{i})-z(x_{i+2}))^{2}$   1   0   36   0   100   81   0   1   256   81   4   0   16     576 13 44.3
150                              
$z(x_{i})-z(x_{i+3})$   3   4   2   14   5   14   -6   -11   -15   -8   15  5      
$(z(x_{i})-z(x_{i+3}))^{2}$   9   16   4   196   25   196   36   121   225   64   1   25       918 12 76.5
200                              
$z(x_{i})-z(x_{i+4})$   7   0   16   9   10   8   -16   -10   -14   -9   6        
$(z(x_{i})-z(x_{i+4}))^{2}$   49   0   256   81   100   64   256   100   196   81   36         1219 11 110.8
250                              
$z(x_{i})-z(x_{i+5})$   3   14   11   14   4   -2   -15   -9   -15   -4          
$(z(x_{i})-z(x_{i+5}))^{2}$   9   196   121   196   16   4   225   81   225   16           1089 10 108.9
300                              
$z(x_{i})-z(x_{i+6})$   17   9   16   8   -6   -1   -14   -10   -10            
$(z(x_{i})-z(x_{i+6}))^{2}$   289   81   256   64   36   1   196   100   100             1123 9 124.8
350                              
$z(x_{i})-z(x_{i+7})$   12   14   10   -2   -5   0   -15   -5              
$(z(x_{i})-z(x_{i+7}))^{2}$   144   196   100   4   25   0   225   25               719 8 89.9

Figure 3.3: Computed Simple Variogram.
/begin{figure}/centerline{/psfig{figure=fig3_3.ps,width=/textwidth}}/end{figure}


next up previous contents
Next: The Variogram Up: Regionalized Variables Previous: Example 3.1: Deposit of   Contents
Rudolf Dutter 2003-03-13