International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
Figure 5 - Example of the As cosimulated map 
  
  
  
  
  
  
  
  
  
30000 199 3 
25000 
20000 . 
T pre J 
1000 / 
{ Variable Cu: 
$0001 } Struct 1: Exponential 
j Range: 450m; Sill: 16552 ppm2 
0e 500 1000 1500 2000 2500 3000 
h (m) 
y (h) 
250000 
200000 . + 
150000| ie 
| / : 
100000 | / 
/ Variable Pb: 
50000 / Struct 1: Nugget effect. Sill: 50000 ppm’ 
| Struct 2: Exponential 
| Range: 300m; Sill: 111309 pom’ 
0 500 1000 1500 2000 
  
  
  
  
  
  
  
  
  
  
  
h (m) 
y (h) 
450000 
400000 : 
350000 . ° 
3000001" A 0$ mid Lii» eA 
250000 wr fs 
200000 . / 
150000| / 
190000). / Variable Zn: 
50000| / Struct 1: Exponential 
à / Range: 750m; Sill: 307762 ppm* 
0 500 1000 1500 2000 2500 3000 
h (m) 
el (h) 
5 . 
Ames PE em 
3 d * . 
A64. 
/ Variable Cd: 
17 Struct 1: Exponential 
/ Range: 340m; Sill: 4.20 ppm’ 
0 T 
0 200 400 600 800 1000 
h (m) 
y (h) 
60000 : 
50000 : : . : 
40000 uU T. 7 
SL LJ 
A 
30000| / 
p^ 
20000 Variable As: 
Struct 1: Nugget effect. Sill: 20000 ppm’ 
10000 Struct 2: Exponential 
Range: 500m; Sill: 23740 ppm’ 
0 7 ; 
0 500 1000 1500 2000 
h (m) 
Figure 6. Variograms of the simulated data 
5.2 Hazard Maps 
One of the principal aims of this study was to delineate the 
areas that need future remediation based on intervention values. 
The Consejeria de Medio Ambiente de la Junta de Andalucía 
(Environmental Agency of the regional government of the 
South of Spain) C.M.A, (C.M.A., 2000) defined for each 
contaminant four different remediation levels: maximum level 
allowed, recommended investigation, compulsory investigation, 
compulsory treatment. Hence joint probabilities of different 
metals to be simulated simultaneously, above or under the 
remediation levels, can originate hazard maps of the region. The 
study area mainly consists of agriculture soils with pH values 
lower than 7. 
Considering the thresholds: z, (maximum level allowed), z, 
(recommended investigation), z; (compulsory investigation) and 
Z4 (compulsory treatment), the following joint probabilities, at a 
given location xo, can be identified with different hazard levels: 
a) Prob  (zca(Xo)Czice ZeulX0)<ZiCu> Zzn(X0)<Zızm 
Zp(Xo)€Zipo, Zas(Xo)<Zias} corresponds to the most clean 
hazard scenario; 
b) Prob {zca(Xo) Zacas ZeulXo) Zacw  Zzn(X0) Zoz» 
ZpXo) Zap»  Zas(Xo) Zaas), corresponds to the 
intermediate clean hazard scenario, meaning that all metals 
at Xo Zp; 
c) Prob {ZcalXo) Zscas Zcu(X0) Zscw Zzn(Xo) Z3zm 
Zp(Xo) Zspp — Zas(Xo) Zsas}, corresponds to the 
intermediate contaminated hazard scenario, meaning that 
all metals at x, are greater or equal to zs; 
d) Prob {zca(Xo) Zacas Zcu(X0) Zacw Zzn(Xo) Zazm 
Zp(Xo) Zapp Zas(Xo) Zaas}, corresponds to the most 
contaminated hazard scenario. 
We can define a joint indicator using the corresponding 
indicator for the different metals: Iz,(Xo) = Iz cq (xo) . 1Z1cu(X0) - 
Izip(Xo) . Iziz&(Xo) - Izias(xo). The joint probability at xo — 
corresponding to scenario i) — can be estimated with the 10 
simulated images: 
10 
prob, (x,)= n (x,,i) L;(xoi) corresponds to L,;(x)) of 
i=l 
simulated image 7. 
Equivalent joint probabilities can be computed for the other 
scenarios: prob,2(Xo), probz3(Xo) and prob,4(xo)- 
Finally, a global hazard map was obtained by classifying each 
pixel in four defined scenarios (Figure 7). The results show that 
approximately 44% of the study area need compulsory 
treatment and 40% of the study area need compulsory 
investigation. 
Since some European legislation impose treatment whenever 
one metal exceeds the highest threshold (compulsory treatment), 
an alternative for scenario iv was conducted: if at least one 
heavy metal exceeds the compulsory treatment threshold the 
pixel is considered to belong to scenario iv, i.e., treatment is 
imposed to that soil, with Prob {zca(Xo)=Zaca OT Zcu(Xo)7Zacu OT 
Zzn(Xo)=Zazn OT Zpp(Xo)=Zapt OT Zas(Xo)=Zaas} - 
In this alternative, scenario iv has higher probabilities of 
occurring in comparison to the other remediation levels. The 
obtained probabilities for this scenario show that approximately 
376 
Internai 
  
72% of 
of the st 
$3 Envii 
Finally, w 
it is possi 
map. An 
environme 
2000). The 
impact lev. 
occupies 
correspond