i| 2004 
nagery 
ions of 
led on 
on. An 
tion of 
tation 
. This 
n this 
fferent 
| angle 
e and 
image 
Radar 
Boston 
es" of 
pplied. 
Works, 
puting 
signal 
IEEE 
achine 
95, A 
etwork 
d land 
ce and 
is RT, 
error 
rations 
irt and 
Press. 
2000, 
lihood 
forest, 
ensing 
/ecchi, 
220, 
DISTRIBUTED GEOSPATIAL INFORMATION SERVICES-ARCHITECTURES, 
STANDARDS, AND RESEARCH ISSUES 
Liping Di 
Laboratory for Advanced Information Technology and Standards (LAITS), George Mason University, 9801 Greenbelt Road, Suite 
316-317, Lanham, MD 20706, USA — ldi@gmu.edu 
Commission II, WG II/3 
KEY WORDS: Geospatial information Services, Distributed Architectures, Standards, Knowledge Systems, Web Service, Grids 
ABSTRACT: 
It is estimated that more than 80% of data that human beings have collected so far are geospatial data. In order for the geospatial data 
to be useful, information has to be extracted from the data and converted to knowledge. However, currently it is very difficult for 
general users to obtain geospatial data and turn them into useful information and knowledge. In order for geospatial information to 
become the mainstream information, the geospatial information systems have to be able to provide the ready-to-use information that 
fits the individual users’ needs. This paper presents the distributed geospatial services which can fully automate data discovery, 
access, and integration steps of the geospatial knowledge discovery process under the interoperable service framework, fully 
automate a rang of geo-computational services at a limited geospatial domain, and greatly facilitate the construction of complex 
geocomputation services and modeling. The paper discusses the service-oriented architecture, web services, Grid services, geospatial 
services, geo-object and geo-tree concepts and their implementation in the service environment, and the interoperable geospatial 
service standards. It also discusses some advanced research issues related to automatic geospatial information extraction and 
knowledge discovery in the web service environment, including automated geospatial data and services discovery, domain 
knowledge-driven intelligent geo-object decomposition, automated geospatial web service chaining, binding, and execution, and 
management of workflows and geospatial models. 
1. INTRODUCTION management and decision-makings. In other words, we are rich 
in geospatial data but poor in up-to-date geospatial information 
Data is a representation subject to interpretation or to which and knowledge. In order to diminish this problem, we need to 
meaning may be assigned. Geospatial data is the data that can develop the semantic processing that automatically derives the 
be associated with locations on Earth. Geospatial data is the information and. knowledge from the geospatial data in the 
dominant form of data in terms of data volume. It is estimated distributed archives. 
that more than 80% of data that human beings have collected so 
far is the geospatial data. Geospatial data is widely used in 2. THE DISTRIBUTED NATURE OF GEOSPATIAL 
many aspects of socio-economic activities, ranging from DATA REPOSOTORIES 
environmental management to military operations. Because of 
importance of geospatial data, huge amounts of resource and Because of the importance of geospatial data, many public and 
money have been invested in collecting, managing, archiving, private organizations have been engaged in collection, 
and distributing geospatial data. The total amount of geospatial management, and utilizations of the data. Typically, individual 
data is approaching to exabytes and continues to grow rapidly. organizations establish their own data centers to manage and 
For example, NASA EOSDIS have already archived more than distribute the data and products to their users. Because of the 
two petabytes of data, with more then three terabytes of new data ownership issues and the large data volumes, it is 
data arriving each day (McDonald et al., 2003). impossible to put all geospatial data into one big data center, 
even if within a single country. In addition, the computational 
Information is the meaning that is currently assigned to data by resources associated with those data centers are naturally 
means of the conventions applied to these data. Knowledge is distributed. 
an organized, integrated collection of facts and generalizations. 
In order for the geospatial data to be useful, information has to 3. STEPS AND ISSUES IN THE PROCESSES OF 
be extracted from the data and converted to knowledge. GEOSPATIAL KNOEWLEDGE DISCOVERY 
The computer data processing in the geospatial knowledge 
discovery includes three consecutive steps: A) Geoquery, B) 
Geodata and information assembly, and C) Geocomputation. 
Geoquery is to locate and obtain data from data repositories. 
The geocomputation is to analyze and simulate the complex 
Earth system using data and information from the Geoquery. 
Geodata and information assembly assembles the data and 
information from data centers based on the needs of 
Because of the complexity of their formations and the multi- 
disciplinary nature of geospatial data, it is difficult to extract 
useful information from them without sophisticated processes. 
The traditional way for geospatial information extraction 
involves in, with help of computers, the trained specialists who 
are the experts on the specialized area. Because of the huge 
volume of the geospatial data and scarcity of the experts, most 
of the geospatial data have never been directly analyzed by 
people after the collection. On the other hand, the society is geocomputation. 
lacking of up-to-date geospatial information and knowledge for 
187