This distribution is a function of *c*, and, to be com-

pletely defined, needs to be known for all values of

such as: suppose that we have no measurement of

concentration of a certain contaminant at *x*, but the

distribution is known, and a threshold value of

c = 8 mg/l is of interest. If *P *[*Z *(*x*) # 8] ' 0.60 ,

then, if a measurement were made at *x*, there is a

60-percent chance of obtaining a value less than or

equal to 8 mg/l. The distribution also may be used

to calculate other probabilities, such as the proba-

bility of obtaining a value in some specified

interval.

(3) An important concept to keep in mind in

all geostatistical applications is the **support **of the

regionalized random variable. The support of *Z(x*)

is the in situ geometric unit represented by an

individual sample. For example, in a soil contami-

nation study, sample Z*(x) *might represent the con-

centration of a contaminant in a vertical soil core

0.1 m in diameter and 1 m in length, and centered

at location *x*. Thus, even though Z*(x) *is defined at

a particular point, it is representative of a volume

of soil. Changing the support of *Z(x*) will usually

change its probability distribution. Therefore, the

observations in a geostatistical analysis should all

have the same support. The method called point,

or punctual, kriging, described in section 2-4, is

designed to predict values of *Z(x) *with the same

support as the sample data.

(4) A concept closely related to support is that

of **estimation block**, which is a geometric unit

larger than the support of a single observation, for

which a single representative value is desired. For

example, in the above soil contamination study, it

may be necessary to estimate the average concen-

tration of the contaminant in a truckload of soil

excavated from a block 6 m long, 6 m wide, and

0.3 m thick. Using a method called block kriging,

also described in section 2-4, the block average can

be predicted based on individual measurements.

(5) Although the distribution of *Z(x) *com-

pletely characterizes *Z(x) *at any particular loca-

tion, this distribution indicates nothing about the

relations among the values of *Z(x) *at different

2-2