Issue 1 [March]

The ENVIRON model estimates environmental impacts (positive, negative) from the introduction and
use of Advanced Communications (AC); Information Society Technologies (IST) in industrial, commercial
and business sectors in Greece. The model estimates effects on output, employment, income,
environment and energy requirements. It is based on the: (i) Leontief Input-Output theory-analysis,
(ii) Introduction of AC/IST and in particular of the Telematics as a new sector into the economic system
of a country, and (iii) Incorporation of pollution emission factors into the system.
The types of AC represented are grouped into six categories: access to information systems, electronic
transactions, robotics and tele-action, tele-working, mobile communications and video facilities.
Industry sectors considered are transport, business and services, public and domestic.
The application of ENVIRON indicates that the introduction of AC into the production process will
result into a 15.8% decrease of energy consumption, a reduction of 14.32-10.14% in SO2, and it will
have strong positive effects on the economic system of Greece especially on profits related to environmental
protection. The model demonstrates the use of the Leontief Input-Output analysis in environmental
impacts analysis matters and policy.

The objective of this study was to determine whether the use of the co-regionalization of the distanceto-
river topographic variable with the soil properties topsoil clay and sand can improve their mapping.
The interpolation techniques: ordinary kriging, kriging combined with regression (two models) and
heterotopic co-kriging were applied to data from 153 observation points. The two models of kriging
combined with regression involve: (a) linear regression of the two soil variables with the distance-toriver
variable on the 153 observation points followed by kriging and (b) summation of the kriged regression
values and kriged regression residuals. For co-kriging 350 additional observations for the distanceto-
river-variable were employed. The distance-to-river data were easily obtained from the map of the
area which was stored in a Geographical Information System (GIS). The performances of the methods
were evaluated and compared using the cross-validation method. The mean error of prediction indicates
reasonably small bias of prediction for the two soil variables by almost all the methods. The mean
square error showed that heterotopic co-kriging produced better estimates of the soil variables than
kriging but there was a clear advantage in using the first model of kriging combined with linear regression
technique. The second model of kriging combined with regression does not show any particular
advantage over the other methods.