278 
we start an exhaustive correlation and dispersional analysis which will 
allow us to pass from the qualitative to the quantitative elements; this 
is an indispensable procedure for determination of "I”. 
/c/ Risk considerations within remote sensing processes 
pro 
These must be carefully analysed because of their potentially rat 
decisive influence upon the process. Several of such restrictions are inc 
enumerated as an example: ¿ ec 
for 
- The aerial photographs and records cannot be always carried out nuin 
by means of the preestablished sensors and captors because of their un 
availability, or they are out-of-date, or not delivered in time. The un 
suitable meteo-conditions, time difference regarding the space records and in 
field-sample determinations, incorrect flight-line, unperforming the estab- pro 
lished circumstances for the sensors, etc., are also different possibilities may 
to hinder the use of desired sensors and captors: to 
- Some other hindering factors are also: partial or total unavail 
ability of the proper space records, late delivery of these records, dif 
impossibility of having the necessary data from the other segments of the tic 
system. (f 0 
mad 
The list may be continued; its items are necessary to know what the 
elements must be replaced under different circumstances. cal 
prc 
In a mathematical formulation of the problem, the risk considera- pai 
tions appear as restrictions; the problem may be treated by mathematical ing 
programming. 
/d/ Design.of models and their finishing at every stage : of 
pre 
- The term of "model” used here has a broad meaning, supposing a pos 
lot of stages that start with a general model which is at the same time a and 
logical study in a preparatory stage and a data handling plan (Appendix 1). the 
This general model is considered as an initial stage and it is to be con- pre 
tinuously improved until its finishing at every following stage: and 
up, 
- The model is oriented to an area and not to different classical dec 
disciplines; better said, it is pluri-interdisciplinary /3/; 
- The proposed systems may have a more general character but con- ger 
cretely, every system is optimized for its particular purpose; are 
in\ 
- The choice of an optimum remote sensing system obviously implies eqi; 
the comparison of different alternative systems for processing and use or 
of data; app 
of 
- Sometimes the choice of an optimum system may be obvious, without as 
involving special research. However, due to numerous regarded categories, cos 
different available types in every category and accountable interactions 
as well, the optimization problem requires a proper mathematical treatment 
as we will see below.