76
A MAP-BASED WEB SERVER FOR THE COLLECTION AND DISTRIBUTION OF ENVIRONMENTAL DATA
G. Guariso, M. Ferrari, D. Macchi
Politecnico di Milano, Facoltà di Como, Como, Italy
KEY WORDS: metadata, data retrieval, world wide web, Internet, client-server architectures.
ABSTRACT
Environmental data constitute a very wide and poorly structured knowledge domain. The only characteristic that they
share is to be associated to a certain geographical object, either a point, or a line, or a polygon. These geographical
objects have in turn known relationships among them (e.g. inclusion, closeness). A map can thus constitute a unique
way for structuring such a wide domain, thus allowing an ordered data storage and retrieval. A web server for collecting
and distributing environmental data has been built on this principle. Its main aim is to facilitate the sharing of information
among all the subjects and institutions operating on a given territory, for planning, management and also educational
purposes. The system adopts the classical client-server architecture and the user can contribute to its growth under the
supervision of an administrator. It has been implemented for the Lombardy region in Northern Italy at the address
http://pc-ambiente. como.polimi. it.
1. INTRODUCTION
Environmental data cover a very wide and heterogeneous
knowledge domain: they refer to different environmental
compartments (air pollution, water quality and quantity,
population dynamics, and so on), they are usually
collected by different agencies at different locations and
time intervals and they are stored in different ways.
Proposals for a standard environmental data format have
been put forward by several institutions (e.g., FGDC,
1994: Library of Congress, 1997; Phelps and Wilensky,
1996; Koschei et al., 1997), but have not been widely
adopted. Environmental databases are growing up in
dimension and complexity so as to make any search
extremely difficult (see, for instance, the US Geophysical
or Climatic Data Centers or the NOAA Environmental
Services Data Directory — http://www.edsim.noaa.gov/
NOAA-Catalogue/). Their production and maintenance
requires an effort in terms of economic and human
resources that cannot be sustained by all institutions.
It is thus of interest to exploit the only characteristic that
environmental data share: to be associated to a certain
geographical object, either a point (for instance, a rain
gauge) or a line (the flow rate of a river), or a polygon (the
population of a municipality). These geographical objects
have in turn some known relationships among them (e.g.
inclusion, closeness), that can also be automatically
managed. A map can thus be used to provide some
structuring to such a wide information domain, which in
turn may allow for an easier data storage and retrieval.
A web server for collecting and distributing environmental
data has been built on this principle. Its main aim is to
facilitate the sharing of information among all the subjects
and institutions operating on a given territory, for planning,
management and also educational purposes.
The storage and retrieval system is based on the classical
client-server architecture, but cannot be considered as a
standard database since it grows up with the participation
of the users, on the client side, and is only supervised by
an administrator, on the server side. Additionally, there is
no specific structure for the data stored in the system,
except their reference to a geographical entity.
After briefly revising some map-based environmental
information systems in Section 2, the paper presents in
the following sections, first the data structures and then
the architecture of the web site. Section 4 gives some
technical detail on the system implementation and Section
5 provides some examples of its use.
2. WEB BASED GIS SYSTEMS
The diffusion of the Internet has significantly changed the
approach to storage and distribution of environmental
data. Problems related to the planning, management and
conservation of the environment are, for their own nature,
distributed in space. Various public authorities and
agencies, as well as private citizens are normally involved
in any such problem. The Internet is thus a natural mean
to share the information, support the debate, create
consensus and thus ease the implementation of solutions.
Unfortunately, the set of data required for a specific
environmental problem is relative unique, in many
instances, and thus it is almost impossible to structure a
classical database that may be useful in different
situations.
Fig. 1 - Berkeley's Digital Environmental Library. The photo
is obtained by clicking the required position on the map
