International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B1. Istanbul 2004
sprawl. These are common problems throughout Europe
and the UTS recognizes that there are clear opportunities
at the European level to develop, share and facilitate the
implementation of appropriate solutions.
In this context the UTS approach is to develop a strong
framework at the European level to provide this coordi-
nated approach and more systematic support to cities. Specif-
ically the UTS focusses on sustainable urban management
and urban transport and proposes the development of an
integrated framework for tackling the different complex
issues and to establish coherent and coordinated environ-
mental policies within towns and cities. The approach pro-
posed is that the largest towns and cities in the EU 25 (the
500 or so towns and cities over 100,000 inhabitants) are re-
quired by law to develop and implement an urban environ-
ment management plan and an environmental management
system to ensure its implementation.
It is clear that the UTS generates entirely new policy driven
information and intelligence needs based on the urban en-
vironmental plan and management system proposed. These
information needs are conceptually specified in terms of a
cycle of evaluation, target setting and monitoring, based
on objective and appropriate data and indicators. The tools
and methods essential to deliver this integrated approach to
urban management create new demands for higher quality
information and intelligence as well as for new concepts of
information management.
As an illustration of these new information needs the fol-
lowing listings provide an overview of requirements at both
EU and local levels:
e EU Level information needs
e — Monitorand guide implementation of Urban The-
matic Strategy
— Monitor Directives enforcement of EU legisla-
tion
— Regional Land Use Monitoring
— EEA “state of the urban environment” report
— European Topic Centre coordinate information
needs
— New EU level environmental indicators
— New EU sustainable urban transport indicators
e Local Level information needs
e — Report on implementation of LA21
— Harmonization of urban environmental data
— Plans based on explicit environmental targets,
* actions and monitoring
— Need for land use, noise, transport, air quality
indicators
— Data and indicators to be applied according to
standards
— Develop data to indicators for reporting at na-
tional and European levels
— Maximize use of urban management tools and
models
320
3 GUS PRODUCTS
As already recalled GMES Urban Services (GUS) isa project
from ESAs GMES Service Element (GSE) aimed to the
consolidation of a product portfolio addressed to meet the
requirements from urban areas. GUS team is currently
made of 11 companies and research teams with the follow-
ing roles: Service providers (Planetek, Eurosense, Hugin,
Indra, SCOT); System Developer (Definiens); Consultants
(ControlWare, UWE); Research Partners (Department of
Electronics, University of Pavia and Environmental Stud-
ies Centre-Vitoria-Gasteiz).
In headlines, GUS portfolio is made of three big compo-
nents:
e UTS block. Information in support of the UTS, made
of different geoinformation layers derived from Earth
Observation satellites combined with ancillary infor-
mation from in-situ measurements including socio-
economic data.
e REG block. A set of products at regional scale fo-
cused in urban areas and spatial planning.
e DEV block. A set of products for cities in developing
countries (collaboration with UN-Hab).
3.1 UTS block
Among the different products addressed by GUS in the
UTS block, we would like here to focus on land use map-
ping. Reason is that this application is extremely more
challenging than land cover mapping. Of course, the two
tasks have something in common, since they are both clas-
sification problems. However, some land use classes are
impossible to obtain from remote sensing data only, and
land use mapping always requires at least a spatial analy-
sis. As a matter of fact, the use of a land parcel may be
different even with the same cover, but may be inferred by
its neighborhood.
The problem of land use mapping in urban and periur-
ban areas has been traditionally approached in two ways.
The first starts from a detailed land cover map obtained
by means of standard or improved classification routines.
Then, spatial reclassification by kernel techniques is ap-
plied to extract from land cover classifications the land use
maps. Alternatively a priori knowledge or external data
sets is considered by a human expert or a knowledge sys-
tem. The second methodology is instead based on a direct
approach, which tries to incorporate into the first classifica-
tion step some kind of spatial (texture) information. Thus,
land use maps are directly extracted.
Considering the approach to land use mapping based on
extracting first the land cover and then land use classes, an
overview of the present methodologies requires first some
considerations on land cover mapping in urban areas, and
then on the techniques used to provide land use from land
cover maps.
Cy o ONY AN nn Mm
"7.
m