You are using an outdated browser that does not fully support the intranda viewer.
As a result, some pages may not be displayed correctly.

We recommend you use one of the following browsers:

Full text

International cooperation and technology transfer
Fras, Mojca Kosmatin

JUNE '93-JUNE'97
Lojovic E.H., Sabic D. and Tretjak A.
All authors: Department of Statistical Geomatics and GIS
Statistical Office of the Republic of Slovenia
Vozarski pot 12; SI-1000 Ljubljana; Slovenia
The ISPRS WG VI/3 and WG IV/3 joint meeting: Bridging the Gap
KEY WORD: Land cover change, Land cover, Land use, Geocoded data layers, Satellite scanned data, centroids of
houses, GIS
In order to estimate the size, location and previous land cover of augmented built-up areas in Slovenia in the period from
June 1993 to June 1997 the data layers of built-up areas and Statistical Land Cover/Land Use GIS - state '93 were used.
The Statistical Land Cover/Land Use GIS was compiled at the Statistical Office of Republic of Slovenia using several
existing georeferenced data layers of which the georeferenced Landsat-TM/93 satellite scanned data present the
thematic layer. That layer was visually interpreted with the minimum mapping unit of 20 hectares. The augmented built-
up areas were obtained from two data layers containing centroids of houses: June 1993 and June 1997. They were
buffered and merged, the common areas were eliminated and the remaining were overlaid onto the compiled Statistical
Land Cover/Land Use GIS of Slovenia - state '93. Thus the land cover categories that turned into built-up areas were
As expected the density of new built houses is the highest around urban centres and most of the new houses were built
on agricultural land. It is worth pointing out the area in Central Slovenia (Savinja valley) which is categoriesed as first
quality agricultural area, where the new built-up areas are nearly evenly distributed over the whole valley. It is important
to stress that the simultaneous use of land cover and built-up data layers enables a quick and simple i.e. visual
identification of the location and extent of this kind of land cover changes. In addition, it can be used also as a correction
tool for the producers of these data layers as some illogical locations of centroids of houses are immediately visually
detected. It is our goal to analyse the same change for the period from 1997 to 2001, using the Statistical Land
Cover/Land Use GIS of Slovenia - state '97 with the minimum mapping unit of 15 hectares with the augmented built-up
areas data layer, which will, in addition to the centroids of houses, include also built-up areas of larger industrial objects,
warehouses, parking places, etc., as well as the railways and the first three levels of roads.
In 1997 the first Statistical Land Cover/Land Use GIS of
Slovenija-state'93 was compiled merging a number of
georeferenced data layers, all from 1993 and covering the
whole Slovenia (2,027,245 ha):
Landsat-TM satellite scanned data,
digitised administrative boundaries,
digitised boundaries of wooded areas,
digitised boundaries of water bodies,
digital elevation model: 100 m by 100 m
centroids of houses,
vectors of railways,
vectors of first level roads.
These data were used for the delineation of the five main
land cover/land use categories (Duhamel, 1995):
1. Wooded land cover,
2. Agricultural land use,
3. Areas under water,
4. Bare rocks, i.e. non-vegetated land cover,
5. Man-made land use, i.e. houses, roads, railways.
In addition, the category "undefined areas" was
introduced, comprising 0.01% of the classified territory.
The three elements of the 5 th land cover category, i.e.
houses, roads and railways, were obtained from three
separate data layers. The first data layer was acquired
from buffered centroids of houses the procedure of which
is described in the next chapter. The central lines of roads'
and railways' data layers were buffered according to the
official width of corridors of roads and railways
respectively. These buffered data layers were merged,
crossings and overlapping areas cleaned and used as
one data layer presenting the "man-made" land use
category. (Sabic, 1998; Lojovic, 1999).
In the Register of Territorial Units the records of only
those houses that have house numbers are collected. The
houses are defined with geographic co-ordinates as: "a
unique identification of every residential or commercial
building". The geographic co-ordinates indicate the centre
of each house and are named centroids. The centroids
are thus points, that do not bear information on the use of
the building or on its area. In order to estimate the area
under every building, including the average area of the
pertaining yard or garden, the centroids were buffered
with a radious of 20 m (Peled, 1993). The selection of a
20 radius was based on a statistical analysis of sampled
individual buildings over all Slovenia, and the areas of
yards, gardens and auxiliary buildings were measured on