ted
ga
lly
ner
ge,
em
[he
the
lent
; of
asic
ycal
age
olor
DIC
T 10
ICS.
LANDUSE AND URBAN DENSITY MAPPING USING REMOTELY
SENSED DATA: THE CASE STUDY OF HAIFA, ISRAEL
Maxim Shoshany
Eran Wechsler
Department of Geography
Bar-Ilan University
Ramat gan 52900, Israel
ISPRS Commission VII / Working Group 9
ABSTRACT
Analysis and mapping of landuses and landcovers represent one of the most dynamic fields of remote
sensing applications (Townshend, 1990). In urban regions, for example, Weber and Hirsh (1992) have
assessed urban life qualities and Gomarasca et al, (1993) monitored longterm changes in the built up
areas. In the rural-urban fringe Jensen and Toll (1982) detected development of residential areas;
Quarmby and Cushnie (1989) monitored changes in the use of agricultural lands. Recent developments of
methodologies for recognizing landuses have moved from the classical multispectral approach to
techniques utilizing spatial and temporal information. Gong and Howarth (1992) for example, developed
frequency-based contextual classification for landuse identification. Jewell (1989) for example, used
temporal changes in reflectance for differentiating between types of landcovers.
The use of remote sensing methodologies for estimating urban densities is well described in the literature.
Most of these works were based on manual interpretation (see review in Lindgreen, 1985). Individual
objects such as building and chimneies were mapped and counted. The results were shown to be of high
accuracy relative to census data. However, manual interpretation requires long tedious processes thus
limiting the size of the regions which could be monitored. Only in few works there was an attempt to
map densities through the use of satellite images and computerized techniques. Results reported by
Forster (1985) and Keersmaecker (1990) have indicated difficulties in achieving good results due to the
relatively low resolution of those images.
The Metropolitan area of Haifa represents one of the most complex spatial composition of landuses
together with massive expansion during the last decade due to the large immigration waves. The
objective of this study was the development of computerized methodologies for differentiating between
commercial / industrial and residential areas, for identifying areas belonging to different urbanization
stages and for mapping different urbanization stages and for mapping different urban densities.
141