Geoscience Reference
In-Depth Information
interaction can initiate undesirable changes in the environmental characteristics. For
instance, when assessing the risk of human health from the environmental pollution,
to make a serious decision, a certain level of the informative description of the
territory is needed:
multi-year observations of concentration of chemical elements with their indi-
cated allocation and characteristics;
data on hydrology of the territory and synoptic characteristics;
assessment of ecological consequences of pollution and their impact on human
health;
characteristics of the state of protective constructions and the level of devel-
opment and technical equipment of services for monitoring, prevention and
meeting extreme situations.
ciency of the systems of risk assessment depends on the form and kind of
the applied procedure of decision making (Potapov et al. 2006; Garsey and
McGlade 2006; George 2002). The most informative is when prediction procedures
are combined with an environmental monitoring regime, which can foresee situa-
tions of decision-making in a real time on the basis of information accumulation
before the decision was made or from analysis of database fragments with no time
reference. Statistical analysis of several events that follow the functioning of the
monitoring system can be carried out by numerous methods whose applicability in
each case is determined by the totality of probabilistic parameters that characterize
the phenomenon under study. However, non-stationary and parametric uncertainty
in situations when each observation requires much effort and expense prompt a
search for new methods of decision-making on the basis of observation data that are
fragmentary both in time and in space.
With development of alternative methods of making statistical decisions, the
problem of
An ef
finding objective estimates of the perimeters of processes taking place in
the environment has been substantiated anew. It is possible to consider and compare
two approaches to this problem: classical approach based on a-priori restricted
number of observations, and successive analysis based on the procedure of step-by-
step decision making. The development of computer technologies facilitates real-
izing both approaches in the form of a single system of making statistical decisions.
The classical and sequential procedures of making statistical decisions by the
Neumann-Pearson (Neyman and Pearson 1933) and Wald (1947) methods were
The notion of risk is closely connected with the notion of uncertainty of pre-
diction of the development of events in the NSS undesirable for humans. By
de
nition of Burgman (2005), risk is the chance of occurrence of an unfavorable
event within some time interval. For the NSS such intervals are measured by
centuries, and the society,
first of all, wants to know the prospects for life
improvement. However, the state of the current environmental science does not
provide a guaranteed solution of this problem, and therefore it is necessary to search
constructive technologies of risk assessment. One of the reliable and constructive
approaches to the solution of this problem is an expansion of the GMNSS
Search WWH ::
Custom Search