Computers, Environment and Urban Systems

27 (2003) 9–32
www.elsevier.com/locate/compenvurbsys

A real-time hydrological model for flood

prediction using GIS and the WWW
W. Al-Sabhan *, M. Mulligan, G.A. Blackburn 1
Department of Geography, Kings College London, University of London, Strand, London, WC2R 2LS, UK

Abstract
The purpose of this paper is to examine the current status of real time hydrological models
used for flood nowcasting and hazard mitigation and indicate how WWW-based systems can
overcome some of the limitations of existing systems. Whilst hydrologically innovative and
robust models are available, they are poorly suited to real time application, are often not well
integrated with spatial datasets such as GIS. Current systems also lack flexibility, customisa-
bility and accessibility by a range of end users. We describe the development of a Web-based
hydrological modelling system that permits integrated handling of real-time rainfall data from
a wireless monitoring network. A spatially distributed GIS-based model is integrated on the
basis of this incoming data, approximating real-time to produce data on catchment hydrology
and runoff. The data can be accessed from any WWW interface, and they can be analysed
online using a number of GIS and numerical functions. We discuss the potential users of such
a system and the requirements for interfacing model output with these users for hydrological
nowcasting and spatial real-time, emergency decision support. Rather than discuss develop-
ments in the modelling of hydrology for flood hazard mitigation, this paper focuses on
developments in interfacing these models with end users.
# 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Geographical Information System; Web-based GIS; Real-time hydrological model; WWW GIS

1. Introduction

Floods are major contributors to personal injury and to property damage, and
can strike with little warning. Problems related to flooding have greatly increased
over recent decades because of population growth and the subsequent development

* Corresponding author. Tel.: +44-207-7848-2632; fax: +44-207-7848-2287.

E-mail address: [email protected] (W. Al-Sabhan), [email protected] (M. Mulligan),
[email protected] (G.A. Blackburn).
1
Present address: Department of Geography, Lancaster University, Lancaster LA1 4YB, UK.

0198-9715/03/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0198-9715(01)00010-2
10 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

of extensive infrastructures in close proximity to rivers. Increased frequency of

extreme rainfall events, characteristics of a changing climate, can also potentially
contribute to this problem. An effective real-time flood modelling and prediction
system could help mitigate the worst effects of flood disasters through the rapid
dissemination of information regarding threatened areas, that is simple maps of
potential flood water distribution. The development of such a system would be
particularly important outside of urban areas where conventional hazard manage-
ment systems do not operate — for example, the large expanses of agricultural land
and pasture which occupy flood plains. There are, a number of legal and institu-
tional difficulties associated with the provision of real-time flood hazard data in an
automated fashion and direct to potential impactees on the ground. However, these
issues fall outside the scope of this paper.
While many watershed modelling software packages are currently available, few
are well integrated within spatial modelling environments (GIS) and are capable of
non-expert implementation. The models require considerable expertise in hydro-
logical data and model application and are unsuitable for real-time application
because of the types of data required and the interactive nature of their applica-
tion. Here we discuss models that are sufficiently simple and robust to allow auto-
matic application and interactive interrogation by end users with little hydrological
expertise.
The application of watershed models requires the efficient management of large
spatial and temporal datasets, which involves data acquisition, storage, and proces-
sing of modelling inputs, as well as the manipulation, reporting, and display of
results. These management requirements are usually met by integrating watershed
simulation models and GISs, generating the capacity to manage large volumes of
data in a common spatial structure. The integrated systems are then further devel-
oped by combining numerous software packages and mathematical programming
systems, and by incorporating a database management system (DBMS) whenever
necessary. For example, the EGIS project by Deckers (1993). Interaction between
the user and the system relies on a graphical user interface (GUI), which may be
developed by a third party such as a university, environmental agency, or commer-
cial vendor. However, these GUIs vary in quality and sometimes contain exclusive
features not found in other similar programs. Another problem is the use of tem-
porally dynamic processes, which are considered a major obstacle when coupling a
GIS with environmental modelling, (Albercht, Jung, & Mann 1997). Some of these
systems are complex and sophisticated, are designed for general purposes, and can-
not be customised by users for applications in specific circumstances (Bundock &
Raper, 1991).
Several researchers (Bennett, 1997; Maidment, 1993; Wilson, 1996) acknowledge
the lack of sophisticated analytical and hydrological modelling capabilities in exist-
ing systems. The systems also need to be adaptable for specific application and user
requirements. Furthermore, existing systems, including simulation models, also lack
some of the essential analysis tools such as dynamic modelling (Van Deursen, 1995),
with the exception of PCRaster (Wesseling, Karssenberg, Burrough, & Van Deursen
1996), needed by watershed applications. Hydrological analysis itself is often
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 11

hampered by limited data of adequate quality (Goonetilleke & Jenkins, 1999) and
lacks real-time data, which are fundamental to the success of any flood forecasting
system. The goal of flood forecasting is to provide a reliable prevention mechanism
to eliminate disasters and reduce the negative consequences of a hazard. However,
this requirement is not often met by GIS and hydrological models. The widespread
and routine use of conventional GISs in important environmental applications has
been hampered by numerous key obstacles (Albercht et al., 1997) such as user
interfacing, data integration, presentation of dynamic processes in GIS and carto-
graphic modelling language (Map Algebra). GIS systems are built to cover a wide
range of applications and are designed to integrate a variety of environmental data,
allowing them to work together in a readily accessible manner. As a result, even
basic GIS functionality for modelling requires relatively complex software and
hardware, resulting in substantial operating costs. Another factor to be considered
is the integration of simulation models, also important tools in environmental
applications of computer technology, which are capable of significantly advancing
the potential of GIS for environmental simulation and understanding. However,
this complex integration requires significant programming effort and data man-
agement (Burrough, 1997). Many model reviews examine the mathematical imple-
mentation of models and their applications (Beven, Kirkby, Schoffield, & Tagg,
1984; Clarke, 1973). Unfortunately, these reviews often do not acknowledge the
seemingly trivial problems that are of significant concern to users and stakeholders.
One such problem lies in the difficulties arising from interfacing the models and the
data; simulation accessibility to interested groups and the general public is another
obstacle to overcome. Rather than giving a full state-of-the-art review of watershed
simulation modelling systems, impossible in only one paper, we have chosen to
present highlights of some of the more recently published literature. We also review
modelling programs with the principal purpose of evaluating their real-time capa-
bilities, interfacing functionality, and public accessibility requirements.
We have two goals in this paper:

1. to review the current status of spatial hydrological models with respect to their
use for (real-time) flood nowcasting; and
2. to introduce a new Web-based real-time model for interactive nowcasting.

The review provides the rationale for the development of an integrated watershed
simulation model with a Web-based interface that benefits users by easing access to
information and by offering technological transparency, platform independence,
visual interaction with data, a multimedia environment, and cost efficiency. Imple-
menting Web-based GISs would provide local communities with Web access to
environmental databases and help them participate in the environmental decisions
that directly affect them. The remainder of the paper is organised as follows: Section
2 provides an overview of spatial simulation models for flood forecasting. Section 3
presents the rationale for the development of WWW-based systems. Section 4 pro-
vides an overview of the methods used and the results obtained and Section 5
concludes the paper.
12 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

2. Review of spatial hydrological models

Hydrological models are defined as mathematical representations of the flow of

water and its constituents on some parts of the land surface or subsurface environ-
ment (Maidment, 1993). Environmental modelling of this kind provides numerous
benefits: (1) it explains the physical world; (2) it provides decision support to
resources and hazard management (Moore, Turner, Wilson, Jenson, & Band 1993);
and (3) it guides experimentation and research for presenting complex ideas in an
accessible manner (Burrough, 1997). Hydrological models can be considered as
stand-alone programs with data loaded via the import and export facilities within
the model itself without the help of any GIS package, or as coupled to a GIS system
through an interface solely designed for that purpose.

2.1. Stand-alone models

The stand-alone programs depend on a particular computing platform or hard-

ware configuration such as the Windows-based Watershed Modelling System (WMS
reference, 1999) and GIBSI (Mailhot, Roussseau, & Massicotte 1997), the UNIX-
based workstations Modular Modelling System (MMS manual, 1998), or the DOS
and UNIX-based PCRaster (Wesseling et al., 1996). They all run from an individual
computer, or from several computers on a local area network that is not necessarily
connected to a global network, and are then made accessible to the public or to
managers in geographically separate locations.
One of the best known and widely used models is HEC-1, from the Hydrologic
Engineering Center of the US Army Corps of Engineers, which provides a variety
of options for simulating precipitation-runoff processes. A new interactive version,
which replaces the command-line-oriented programs HEC-1 (flood hydrograph
model) is HEC-HMS (Hydrologic Modelling System). This stand-alone rainfall-
runoff modelling program has been implemented on both UNIX (Sun Solaris) and
Windows NT/95/98. It has a GUI, integrated hydrological analysis components,
data storage and management capabilities, and graphics and reporting facilities.
Furthermore, a significant amount of data must be gathered to perform hydro-
logical analysis using HEC-1 (Nelson, Jones, & Miller, 1994). HMS is designed to
be independent of any commercial GIS, so it has no explicit ties to any of them.
Another program that is well known and can be used as stand-alone is the WMS
(Nelson, Jones, & Jorgeson, 1995). It is a graphically based, comprehensive
hydrological modelling environment that addresses the requirements of rainfall-
runoff computer simulations (DeBarry et al., 1999). If it is used as a stand-alone
application, data can be imported or exported to or from a GIS package through a
number of popular files formats. For example, a DEM can be imported from
ARC/INFO, or GRASS and then used in WMS for further analysis. The
last example is the TOPMODEL, which is classified as a distributed to semi-
distributed model, and a number of papers and applications of this model have
been presented (Beven, 1997; Obled, Wendling, & Beven 1994). It is not intended
to be a modelling package, but rather a set of conceptual tools that can simulate
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 13

the hydrological behaviour of a watershed (Beven, Lamb, Quinn, Romanowicz, &

Freer 1995). Because it has been adapted and applied to suit many modellers’
needs, a standard version has not been developed. Some of these various imple-
mentations are listed in the TOPMODEL manual available at http://www.es.lancs.
ac.uk/hfdg/topmodel.html. Data requirements for TOPMODEL are dependant
upon the version and the application used. The primary drawbacks of using this
model are the lack of a standard version and the preparation of data for each
application that requires a long and demanding period of data analysis (Romano-
wicz, Beven, & Moore, 1993).

2.2. GIS-based hydrological modelling systems

An alternative to using models as stand-alone, is to couple the model with a GIS

package. There are, however, specific problems in GIS-based hydrological model-
ling, but before we discuss them, we briefly need to address some impediments of
current GIS systems. We can identify four principal areas that impede GIS use:
complexity, interfacing, customisation, and platform dependency. At the present
time, GIS systems are large, complex, and difficult to use and have been severely
criticised as being elitist (Pickles, 1995). The hardware, software, and data can be
highly expensive for most individuals; similarly, the training required to use most
GIS packages successfully is extensive. Over time, GIS software has become more
versatile and general, and different user interfaces and useful models have been
added, all increasing the complexity of the interface. In its early development, the
GIS industry focused on system functionality and data handling. However, it soon
became apparent that the evolution of GISs could not continue without developing
a better user interface (Raper, 1991). Several vendors have developed new interfaces,
with offerings such as ARCVIEW from ARC/INFO and others. Whilst these prod-
ucts represent important advances, several modellers were reluctant to use a GIS in
modelling because of the complexity of developing an efficient integration technique
to support the modelling effort (Stuart & Stocks, 1993).
Customisability is the term used to describe the process by which GIS software is
tailored and modified to satisfy corporate, departmental, and user requirements
(Bundock & Raper, 1991). Each organisation requires a system that is in many ways
unique to its own structure with consideration to its specific problems, requirements,
goals and objectives. At present, most GIS and simulation software customisation
processes are time consuming, require technical expertise in many languages, are
expensive, and very often produce poor results (Raper & Bundock, 1993). Schultz
(1993) discusses the use of GISs with increasing complexity in hydrology and indi-
cates that some necessary hydrological routines are missing from most GISs and
new routines cannot be introduced directly by the user. The last point that is
important to consider is that many of these hydrological models and GISs do not
support multiple platforms. Most systems depend on a specific computer platform
or hardware configuration. For example, WMS runs on a Windows platform (WMS
reference, 1999) while MMS runs under Unix environment (MMS manual, 1998);
Arc/Info, the most popular GIS software, runs on a UNIX platform and the
14 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

Windows-based version has been released recently. Thus, platform dependence is

one of the issues that restrict users from data and software sharing.
The approaches for integration of environmental models with GISs have been
extensively discussed and analysed by many researchers (Burrough, 1996; Good-
child, Parks, & Steyaert 1993; Karimi & Houston, 1997; Maidment, 1996; Peuquet,
Davis, & Cuddy 1993). Similarly, the integration of GISs and hydrological models is
well documented in the literature (Maidment, 1993; Romanowicz et al., 1993; Sui &
Maggio, 1999), thus it is not our intent to discuss these in great detail in this paper.
However, we will briefly discuss the two methods available for coupling GIS and
hydrological models: loosely coupled and tightly coupled (Batty & Xie, 1994; Stuart
& Stocks, 1993). In loosely coupled models a GIS is linked to an external model
where operations are performed in languages such as C or Fortran, which are most
suited to this kind of mathematical calculation. To use the external model from the
GIS, a call to the model is placed to calculate certain parameter values and then
store the values back in the GIS database. This method usually involves a standard
GIS package and a hydrological model such as HEC-2 (Djokic, Beavers, & Desha-
kulakarni, 1994). It is stressed by Albrecht et al. (1997), Burrough (1997) and Frust,
Gristmair, and Nachtnebel (1993) that these couplings provide access to incompatible
analytical tools and ‘‘the difference in the data models and in the way relationships
between variables are handled in GIS and hydrological model’’ (Maidment, 1993)
soon becomes apparent. This mismatch also affects the quality, cost, and benefits of
the modelling results (Burrough, 1997) because modellers and GIS engineers have
very different conceptual views of the real world and approaches to their disciplines.
This type of integration requires a number of programs that exchange data from one
application to another and possibly a DBMS and/or number of transfer files. A
disadvantage of this approach is that there is no common graphical interface and the
data exchange and conversion between the GIS and the hydrological model can be
very cumbersome. Loose coupling may also involve considerable work in changing
data formats and data structure, particularly if the model has been obtained from
another source (Burrough, 1997).
Tightly coupled models are developed entirely within a GIS environment through
the use of a macro language such as AML (ESRI Arc Macro Language). However,
this type of programming is often not able to implement complex applications and
does not support the same capabilities of procedural programming languages.
Unlike loose coupling, tight coupling does not require file conversion or editing;
however, it is a complex process and requires a great deal of programming and data
management (Burrough, 1997) plus a customised menu-driven user interface for
display (Karimi & Houston, 1997). Thus, tightly coupled models require the con-
struction of an appropriate interface that can interact with the data structures of the
GIS system. Karimi and Houston (1997) stress that both the loosely and tightly
coupled methods have inherent limitations for modelling.
While GISs provide powerful tools for spatial analysis, their capabilities for spa-
tio-temporal analysis are limited (Worboys, 1995). Traditional simulation models
are effective in complex and dynamic situations, but they often lack the intuitive
visualisation and spatial analysis functions that a GIS offers. The increasing and
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 15

ongoing literature on the integration of GISs with hydrological models confirms

the recognition of these reciprocal benefits (Maidment, 1993; Moore, 1996). How-
ever, the limitations of current GIS applications for environmental modelling
become apparent because of incompatibility between the types of models and GIS
and the type of data that are associated with them, as discussed by Burrough
(1996), Goodchild et al. (1993). When integrating the WMS with a GIS package
through the use of scripts (i.e. AML and Avenue), developers are restricted to the
architecture of the GIS and users, including hydrologists and engineers, are forced
to learn the GIS operating environment in order to create hydrological models.
Thus GIS specialists often perform the job of the hydrological engineers (DeBarry
et al., 1999). Also when integrating models like MODFLOW (McDonald & Har-
baugh, 1988) and the MMS with a GIS, the outcome is the development of two
different processes (DeBarry et al., 1999; Karimi & Houston, 1997): the pre-
processors and post-processors which are developed in order to facilitate the inte-
gration of software. The pre-processors are tools that prepare, analyse and input
spatial and time series data for use in a model. The post-processors are tools that
perform the analyses of the model results. These include a variety of statistical,
and graphical tools, which can be used to assist in making decision, and a user-
developed interface to display and analyse results. However, these processors are
also designed to create the often complex and cumbersome input file requirement
and to format model output (Karimi & Houston, 1997), including data preparation
of spatial and time-series, analysis, and model output file for display in a particular
graphic environment. Finally many hydrological processes are time-dependent, and
hydrological models generally require time-series input data and generate time-
series output. Current GISs are not equipped to store or manipulate time-series
data; therefore, the ability to store, retrieve, and perform operations on time-
series data is crucial for implementing serious hydrological modelling within a GIS
(Maidment, 1993).

3. Requirements of the hydrological models of the future

3.1. Real-time

Real-time data refers to spatial and non-spatial data that becomes available to the
real-time GIS, either at fixed time intervals or after the completion of certain events
such as the arrival of data at a desired destination. Conventional GIS models such
as data modelling, data management, and software design and engineering do not
allow current GIS systems to meet the requirements of real-time applications effec-
tively (Karimi & Chapman, 1997). The key requirement of real-time flood forecast-
ing, however, is based on continuous in situ measurements of rainfall to improve
the accuracy of model forecasts. While there have been significant advances in the
accuracy of quantitative measurements and the forecasting of rainfall using weather
radar and more prolific and sophisticated rain gauge networks, extensive research
efforts are still required to develop systems that incorporate real-time data with a
16 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

GIS application so that real-time data can be obtained with sufficient timeliness and
rapid nowcasts can be produced. Karimi and Blais (1997) provide a comparison
between current and future GIS systems, arguing that interfacing of GIS with
external systems for real-time processing is nearly impossible. GIS and hydrological
models, however, lack a direct connection with external sensor and devices, resulting
in limited access to real-time data. These deficiencies can lead to hard-coding of data
directly into the system, making updating at existing data particularly difficult.
Karimi and Chapman (1997) suggested that real-time GISs must contain algorithms
that allow fast responses within time constraints for real-time applications. Because
current GIS algorithms do not take these constraints into consideration, they cannot
be used for real-time applications.
Despite advances in computing power and programming languages, FORTRAN
remains the language of modellers (TOPMODEL, HSPF, HEC). Fortran and other
traditional programming technologies are not suitable for the representation of
dynamic geographical systems because (adapted from Bennett, 1997): (1) early or
static binding inhibits the representation of processes that change through time, (i.e.
where one process is replaced by another (different) process); (2) data representation
and management mechanisms are not provided for the development of complex
objects; (3) user defined spatial relations are difficult to capture; and (4) models
developed by traditional programming languages are difficult to extend or modify.
Object-oriented programming techniques provide one means to overcome these
limitations. Raper and Livingstone (1995) demonstrate the use of object-oriented
concepts in the representation of spatial data. Many of the hydrological models
involving simulations and optimisation require significant computer resources
and run for several hours (Walker, 1991), which is unacceptable in real-time
applications.
Currently, there are three basic systems for providing precipitation measurements
that can be used for real-time flood forecasting: first, the conventional telemetry-
based rainfall gauges, the most commonly and widely used, are typically connected
to a base station by telephone lines, VHF/UHF radio, metro-burst telemetry, or
satellite (Latkovich & Leavesley, 1993). A second precipitation measurement system
is weather radar, which provides data on spatial patterns of rainfall as opposed to
point measurements indicated by rain gauges. Meteorological radar is not used as
often because it is expensive and requires sophisticated technical and engineering
support. A review of radar-based precipitation measurement for hydrological
application can be found in James, Robinson, and Bell (1993), and Mimikou and
Baltas (1996). Third measurement tool is satellite monitoring, which estimates rain-
fall by observing radiation signals reflected or emitted from the ground and atmos-
phere. Real-time estimates are possible and, once the satellite is in orbit, the process
is relatively inexpensive. This method provides large area coverage, but also requires
calibration with ground-measured data.
Most data acquisition systems are interfaced to a data transmission system such as
telemetry or wireless radio using standard interface connections such as RS-232
ports. Data can be transmitted from a remote site to a central base station through
several available communication systems. The choice of one communication system
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 17

over another depends on a variety of matters: the size of the watershed, the time of
data transmission, and the costs. The existing types of real-time flood forecasting
systems, which generate the most reliable forecasts, are sophisticated, use very
complex data input and are expensive (Feldman, 1994). For a GIS to be effectively
applied in management decision-making requires reliable, up-to-date data sources
be utilised in building and maintaining the system. Although, flood forecasting sys-
tems exist in many countries, deployed in many projects, real-time flood forecasting
systems are still under significant research and development.

3.2. Interfacing the user

During the development of a modelling system, the most significant aspect of

implementation is an appropriate user interface, for it determines the interaction
between the computer system and the user (Dodson, 1993). This interface permits
the application of a variety of tools regulating both the visualisation and analysis
of the spatial distribution of model parameters as well as the simulated state vari-
ables set at a variety of spatial and temporal scales. The quality of the user interface
connecting the GIS has become an essential contributing factor motivating people to
use GIS as a means of handling spatial data and for determining policies. Frank
(1993) agrees, stating that the user interface is the most important single part of the
system.
In the USA, significant design initiatives have furthered GIS user interface design
such as the National Center for Geographic Information and Analysis (NCGIA;
Kuhn, Willauer, Mark, & Frank, 1992). Albercht et al. (1997) stated that the fin-
ished product of a modelling system must include a graphical user interface that is
simple to use. Knill (1993) stresses the need for improvements in user-friendliness,
including visualisation and results. Some authors focus on the need to develop an
adaptive user interface, which provides multiple, customised interfaces for each
class of users. For example, hydrologists can be presented with a more technical
interface than casual users. Kingston, Carver, Evans, and Turton (2000) also stress
the need for the development of an interface that can be modified to match the skill
levels of the user.
GIS and simulation models are used most frequently in professional and academic
endeavors. Therefore, the developments in watershed model integration with GISs
are actually the result of the development of interfaces that facilitate the creation of
watershed model input data sets (DeBarry et al., 1999). Models such as HEC-1, TR-
20, SWMM, and others which lack good interface design require intensive data
development for input which is often very labour intensive and time consuming
(DeBarry et al., 1999). Even programs such as HEC-HMS, which are designed with
an interactive GUI to provide a convenient means for entering data and displaying
results, the use of the GUI is cumbersome when it is desired to execute a series of
model applications with alternative sets of parameters (HEC, 1996). Because inter-
facing is the most important part of hydrological model development, the need for
improvement in user-friendliness, and flexibility is definitely needed in order to
facilitate the use of such systems.
18 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

3.3. Accessibility

Environmental information is usually held in government, academic, and com-

mercial water institutions. Hydrological modelling tends to remain in the domain of
the model developer and to be applied within a consulting framework. The models
are inaccessible to decision makers who are not specialist modellers (Taylor,
Cameron, & Haines, 1998) and appear to be designed for experts and professionals
for use as in-house tools. While data are currently available for public use through
the Internet by some institutions and governmental agencies, many of these institu-
tions, however, provide the data in varying formats and use complicated relational
databases (Dai, Evans, & Shank, 1997). However, some, like the United States
Environmental Protection Agency (USEPA) and United States Geological Survey
(USGS), have made significant progress toward providing public access, in the form
of real-time data (Stewart, 1999), to various environmental databases. Unfortu-
nately, effective use of these data requires specialised software not available to every
user. Thus, providing public access and analysis to various environmental databases
could help local communities and researchers to participate effectively in environ-
mental decisions that directly affect people’s lives.
Public participation and involvement are important components in the imple-
mentation of a water resources project. Wood, Gooch, Pronovost, and Noonan (1985)
stress the importance of public consultation and participation in flood management
planning. Recently, geographers have begun to consider the practical and societal
impacts of using GISs to support public participation (NCGIA, 1995, 1996). In
addition, it has long been recognized that GISs provide the user with a flexible frame-
work for the development of GIS-based Spatial Decision Support Systems (SDSSs;
Clarke, 1990). A GIS is often designed for spatial decision support, lacking the capa-
city for collaborative spatial decision-making (Jankowski, Nyerges, Smith, Moore, &
Horvath, 1997). However, in the future, a GIS must support SDSSs in collaborative
environments (Karimi & Blais, 1997) designed for sharing, executing, and comparing
model results, especially when the decision makers and/or stakeholders are from
geographically separate locations (Carver, Frysinger, & Reitsma, 1996).
Despite the progress in flood forecasting systems, some researchers in the UK
(Parker, Fordham, Tunstall, & Ketteridge, 1995) have been critical of the Environ-
mental Agency for not giving more attention to warning and dissemination pro-
cesses. More work is needed on increasing community involvement in establishing
local flood warning schemes (Haggett, 1998; Parker et al., 1995).
Finally, data analysis is not only useful for professionals and decision makers but
also for a public that demands openness, as is reflected in legislation such as the
Council of European Communities Directive (1990) and the US Freedom of Infor-
mation Act (1965), which specify that all information must be available to anyone
who wants it. Data accessibility and participation are one of the issues that need to
be considered in the development of a successful flood forecasting systems. There-
fore more consideration needs to be undertaken in order to provide a better simu-
lation and warning system not only for flood forecasting but various environmental
management systems.
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 19

Eight key conclusions can be drawn from this review of current hydrological
modelling. These key points include:
1. Interfacing is difficult. It requires tedious, and complex development and typi-
cally presents the user with difficult to use and limited visualisation of complex
information.
2. Knowledge of GIS is required. GIS and simulation systems are complex, diffi-
cult to use and expensive computer programs, which limit the number of users.
A detailed knowledge of the underlying GIS architecture is required.
3. Platform dependency. Conventional hydrological modelling systems are plat-
form-dependent because every program runs on a specific platform. This
results in limiting the number of targeted users that can access the system.
4. Computer and programming knowledge is necessary. Technical expertise in
multiple languages and subsystems are required (e.g. Databases definition lan-
guage, macro language, C and Fortran, operating system commands).
5. Customisation of models is very difficult. The customisation required to imple-
ment a GIS can run many times the cost of the hardware and software. With
each upgrade, the custom software must be updated, recompiled and some-
times rewritten from the ground up.
6. Limited accessibility for data and analysis must be rectified. Models are still
mostly confined to research laboratories, water and academic institutions and
not widely used in the public at large.
7. Limited collaboration. There is a growing need towards consortium-based pro-
jects, in which projects are, performed by teams of collaborating modellers
from different geographical locations.
8. Real-time data acquisition and communication are costly. The methods and
equipment of data acquisition and communications used are costly.

4. Aims for developing a novel web-based GIS hydrological model

The Internet and the Web are exceptionally important tools, which can potentially
contribute to the achievement of the goals of watershed analysis. The benefits and the
advantages that the Internet and the Web offer to watershed analysis are openness, a
user-friendly interface, interactivity, flexibility, and fast communication. It is rela-
tively cheap and therefore gives the general public access to a variety of both GIS
systems and data of varying degrees of sophistication. This direct access, as a means
of allowing wider involvement and participation in environmental decision-making,
is an important prerequisite of watershed management. This accessibility occurred
because of user interface familiarity, portability across many computer platforms and
capability to display different multimedia and hypermedia formats. Additional bene-
fit that is offered by Java programming language is the development of a sophisticated
user-friendly interface, which can interact with data meaningfully and efficiently.
Users can easily interact with the contents of the database or run simulation models,
thus creating their own analysis results. These results being spatial or non-spatial
20 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

data, hydrographs, still or animated images are processed in a Web server and then
displayed immediately in the client machine from any computer in the world. This fast
communication via the Internet is probably the fastest and the most economic media.
Therefore, by integrating a Web-based system with wireless monitoring network,
real-time data can be obtained and disseminated instantly. These important features
and tools that enhance and improve watershed analysis are available and therefore it
becomes a question of using them efficiently and professionally.
The promising step in the development of Web-based GIS systems has been the
implementation of the Java programming language (Arnold & Gosling, 1996). Java
provides tools for creating graphic, networked, and database applications that are
essential for building watershed management. Java features like object-orientation,
GUI, platform neutrality, security, image processing, and multi-threading are useful
for building such applications (Alsabhan, Mulligan, & Blackburn, 1999). The prop-
osed system was conceived through a general need for real-time data acquisition,
data management, a user-friendly-interface, and accessibility of the data and appli-
cations. The primary aim was to develop an integrated system that would overcome
problems related to the above requirements. The system was designed to act as a
means of archiving and for updating hydrological data as well as for using data in
a rapid decision support system. The system has been developed with the recogni-
tion that several user requirements are currently implemented within watershed
simulation models. Therefore, the aims of the project are:
1. Real-time data access and analysis. One of the aims is to provide data accessi-
bility and analysis by permitting access to real-time data through the Web.
2. On-line watershed simulation and analysis. A detailed implementation of
hydrological analysis should be undertaken in order to provide a decision
support system.
3. Development of a user-friendly interface. A well-designed interface should be
implemented to facilitate the use of the system so that users can focus on the
data, not the interface.
4. Platform-independence. An additional aim is to provide data accessibility and
analysis regardless of the computing platform so that whoever has access to the
Internet can run the programs.
5. Inter-departmental collaboration. A system should be published on a project
Web site, which allows users and stakeholders to share data, as well as soft-
ware as means for coordinating activities between users from different geo-
graphic locations.
6. Cost-effectiveness. One of the aims is to use cost-effective software development
and inexpensive equipments.

5. Methodology and results

The approach taken by the authors has been to develop a system that is practical
and can be applied to a wide variety of watershed scenarios where rainfall data input
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 21

is relayed to the system in real-time. This paper is not concerned with the hydro-
logical model used, but rather with the provision of a methodology for a rapid, easy-
to-use, and cost-effective means for implementing watershed simulation models.

5.1. Data and DBMSs

The data capture requirement is two-fold (Worboys, 1995). First, it must provide
the physical devices, such as automatic loggers (e.g. climatic and hydrological data)
and field computers for capturing data external to the system and for writing to the
database. Second, software must be provided for converting data to structures
compatible with the data model of the database, and for checking the validity and
integrity of data before entry into the system. The physical devices used for captur-
ing data in this particular project are automatic loggers, wireless modems, and NT
servers. The data input procedures in this project use rain gauges to measure rain-
fall. Using a wireless modem, input data are transmitted from the field to a second
wireless modem connected to an NT server that stores the data. These data are then
downloaded in near real-time by File Transfer Protocol (FTP) and stored in a
database, which is immediately accessible to the GIS model for further analysis.
Fig. 1 illustrates the project architecture.
A Java-based model is then used to process the data. The primary advantage of
this approach is that it provides the end user with real-time data. During the plan-
ning phase of this project, we began our first version with a database connectivity

Fig. 1. Project architecture.

22 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

that uses the JDBC approach (Alsabhan, Mulligan, & Blackburn, 2000). However,
the JDBC has a number of shortcomings, the most significant being slow perfor-
mance. Therefore, we upgraded the system using a server-side Java application
(servlet). When the Web page is accessed by a user, the servlet is executed on the
Web server. Unlike applets, servlets run on the server, so they can be large and can
perform a variety of tasks without taxing machine resources or network perfor-
mance. Database connectivity using this approach provides a tool that is powerful,
flexible, and easy to use. Java provides many more capabilities and enhances the
functionality of Web-based data access. A Web data access application, however,
interacts with a database and uses the Web as a means of connection with a Web
browser or Java client on the front-end as a common Web server interface. With
associated libraries such as the Abstract Window Toolkit (AWT; Geary & McClel-
lan, 1997), Swing, which is a subset of the Java Foundation Classes (Gutz, 1998) and
threading (Oaks & Wang, 1999), Java provides the developer with all the necessary
assistance for the development of user- friendly and cross-platform GIS which lead
to performance optimisation.

6. The dynamic model

Dynamic models are difficult to run in most GIS because GIS have been designed
for querying and maintaining a static database with static phenomena. Standard
GIS do not explicitly allow dynamic phenomena to be stored and analysed nor do
they provide efficient facilities for iteration through time (Wesseling et al., 1996).
We have developed our own hydrological model that provides a quantitative
description and understanding of hydrological processes. The details and operation
of this model will be provided in future publications as this is not within the scope of
the present paper. The dynamic modelling unit is a batch program that runs con-
tinuously in the system background. The main purpose of this unit is to process the
incoming data by applying hydrological analysis and then to produce the desired
results. The results of this modelling unit are non-spatial (time series) and spatial
(raster images) of different variables (rainfall, soil moisture, etc.). The program was
designed to store the results in specific directories so that the data loader program
can determine when the program was executed. After the data have been processed
by the modelling unit and the results have been stored in their specified locations,
the database loader program uploads the file automatically to the database. Fig. 2
illustrates the system architecture.

6.1. The map viewer

The map viewer is a Java applet designed for map display. The first page of the
model is the map viewer, which incorporates a hybrid map-display system. Aside
from displaying the two image formats most commonly supported by Java and
browsers, GIF and JPEG, non-standard data files such as large data files held on the
server (the DEM for example) can also be handled by Java. To load a DEM image
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 23

Fig. 2. System architecture.

and the results obtained by applying the hydrological functions (e.g. slope, aspect,
etc.), a rendering algorithm is developed which maps between the DEM elevation
data or the function results and a colour palette and then displays the image ‘on the
fly’. This client-side program is designed in order to reduce network traffic and
processing burden on the server when displaying large files such as DEMs. Once the
DEM is displayed, users can select from the different buttons allowing them to
adjust the map view. To adjust the map view, zoom-in, zoom-out, and pan are
supported. The system is also capable of performing animation, permitting users to
see more than one image at a time in the same applet frame, thereby allowing a view
of the changes of a set of images selected by the user. Before the animation applet
can run, a Java application was designed that run at fixed-time intervals and walks
through specific directories (rainfall, soil moisture, etc.) to convert the specific file
formats to GIF. This application then stores each GIF in its predefined directory so
the animation applet can process when it is called. Additionally, the applet generates
a legend for a selected map. It also allows users to obtain some analytical statistics
on maps to obtain minimum, maximum, and average selected cell values within
user-specified windows, or the whole image can also be chosen to refine the analy-
tical process. The mechanics of the access and the analytical process are available to
anyone on the Web who has authorization.

6.2. Analysis routines

The objectives of the data analysis program are to allow the user to set queries and
to retrieve useful information to satisfy the specific requirements of decision-makers,
and an important function of the analysis is the ability to predict what will occur at
a location, at another point in time, and under certain conditions. The most impor-
tant analytical process of the GIS is the provision of capabilities for spatial analysis
functions that are responsible for the manipulation and analysis of the spatial data.
Currently, the analytical capabilities of GISs related to the structure of the database
(raster or vector) are used, and the proposed prototype uses the raster GIS struc-
ture because that raster family is determined to have greater analytical power. This
24 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

system provides the user with two kinds of analytical capabilities as introduced in
the following sections: time series analysis and spatial analysis routines.

6.3. Time series analysis

The time series tool is a Java applet that interacts with non-spatial data in the
database. To query the time-series database, the user first has to click the database
option from the main applet page. An easy-to-use form is presented that is designed
to be user-friendly. Then, the user can query the database by specifying the time
period and the function (Max, Min, Ave) in order to find, for example, the average
rainfall within a specified period of time. The resulting data indicates the highest/
lowest/average level of any variable within a period of time, and some other infor-
mation. Or the user can specify several variables and produce a list that includes the
date(s), time(s), and value(s) of the time-series data in question. Additionally,
the applet supports a selection mode. The form also offers options that modulate the
interval and time step of data presented, enabling transmission from the server of a
more suitable representation of the time series for the user’s purposes. From the
main applet page, the graph option can be selected to visualise the time-series data.
Depending on user-selected parameters, the graph applet viewer shows changes in
rainfall, temperature, and other variables over daily, monthly, and yearly periods as
they occur. The applet provides the user with the option to view any of these chan-
ges as either line or bar graphs. Fig. 3 shows rainfall time series graph.

Fig. 3. Time series graph page.

 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 25

6.3.1. Spatial analysis routines

The spatial database interface offers the user more flexibility by providing a series
of drop-down lists to choose from. The user has to select a query (What, When,
Where is, or Which), a function (Min or Max), and variable (Rainfall, Precipitation,
Soil Moisture, Runoff, Recharge), and the system searches the corresponding table
to produce results according to the query type. To provide the most flexible and
easy-to-use query interface possible, we have built in a series of questions. The user
can select from a drop-down list for the query type, function, variable, and various
additional parameters. From these variables, a query can be constructed to satisfy
given predicates or conditions, that is, to define what return row of the database
should be projected by specifying conditional Boolean operators that search
expressions identifying the tuples to be retrieved by the query. For example:
What is the lowest soil moisture value that occurs in a specified period of time?
When is the highest rainfall value that occurred between two specified periods of
times?
Where in the catchment is the lowest rainfall value that occurred between two
specified periods of times?
Which (where and when) is the rainfall value that is less than a specific value
AND greater than a specific value?
Typical queries specifying a number of variables can thus be accommodated, and
the data is presented in a result frame. Fig. 4 shows the spatial interface.

Fig. 4. Database interface.

26 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

The user can choose from a variety of possible analysis methods, including terrain
analysis such as slope (Fig. 5) and aspect. More complex analyses are also provided,
such as accumulated flux, the process of calculating the accumulated amount of
water and material that flows over a topological network operating over local drain
direction (LDD) maps. Accumulated flux analysis produces a map displaying grid
cells indicating the amount of water or materials that traversed the cells on their way
to the outlet (Fig. 6). Fig. 7 shows the analysis interface.

7. Conclusion

The availability of GISs via the Web is becoming a reality in many fields (Doyle,
Dodge, & Smith, 1998). Therefore, the intriguing question of ‘‘Web GIS: Toy vs.
Tool?’’ (Thoen, 1995) has been, in just a few years, answered by the GIS community
loud and clear. Many non-GIS specialists are beginning to use Web GIS and many
major institutions are moving GIS products and data to the Web.
Real-time data acquisition is increasingly important in many fields, particularly in
hydrology. The acquisition of data is becoming more necessary each day, but
retrieval of that data can be quite expensive. This paper discusses the development of
a real-time Web-based data access system for watershed analysis using cost-effective
tools. A Web-based interface surpasses the limitation of a traditional decision

Fig. 5. Slope result of the DEM in Fig. 4.

 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 27

Fig. 6. Flow accumulation map.

support system that stores data, model, and user interface on a single computer
(Bhargava & Tettlebach, 1997). It is a convenient system and can be applied to a
wide variety of vital applications necessary for the public good.
One of the advantages of the Web is that it generates information presentation on
demand (Kutschera, Schimak, & Humer 1996). With the inclusion of Java for spa-
tial data navigation and temporal data visualisation, a system can achieve flexible
user interaction without sacrificing efficiency (Taylor, Ackland, Walker, & Jackson
1997). Another significant advantage of the Web is that access to it is relatively
inexpensive and stakeholders can form collaborative contacts. A range of innovative
technologies are being developed that offer different ways of modelling over
the Internet. Pieplow (1998) has demonstrated that several other geographical
information technologies are capable of delivering visual, audio, and multimedia
presentations in support of traditional public participation. Doyle et al. (1998) has
shown the potential use of Web-based mapping and virtual reality technologies for
environmental modelling.
Our project exploits recent networking technology to permit Internet users to
obtain delivery of the desired flood-prediction and management data in real-time and
provides the tools to analyse these data. Using the Web to develop data access
and projected analysis for real-time watershed analysis offers tremendous potential.
Not only to the benefit of environmental users, but also for other disciplines who
need to locate and analyse data. We believe that a Web-to-hydrological model
28 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

Fig. 7. System analysis interface (DEM displayed).

interface can play a beneficial role in real-time watershed modelling, as a valuable

research and educational medium (Parson, 1999) and as a link bringing stakeholders
together for collective decisions.

References

Albercht, J., Jung, S., & Mann, S. (1997). VGIS: a GIS shell for the conceptual design of environmental
models. In Z. Kemp, Innovations in GIS 4, fourth national conference on GIS research UK (GISRUK)
(pp. 154–165). London, UK: Taylor & Francis.
Alsabhan, W., Mulligan, M., & Blackburn, A. (1999). Deploying the client-side of the web for watershed
modelling. In Proceedings of geosolutions: integrating our world, GIS ’99, Vancouver, Canada, March
1–4 (pp. 63–67).
Alsabhan, W., Mulligan, M., & Blackburn, A. (2000). Deploying the client-side of the web for watershed
modelling. In Proceedings of GIS ’2000, Toronto, Canada, March 13–16, conference CD-ROM.
Arnold, K., & Gosling, J. (1996). The java programming language. Java Series, Addison-Wesley Pub. Co.,
2nd edition, December. Mountain View, Califorina.
Batty, M., & Xie, Y. (1994). Modelling inside GIS: part 1. Model structures, exploratory spatial data
analysis and aggregation. International Journal of Geographic Information Systems, 8(3), 291–307.
Bennett, D. A. (1997). A framework for the integration of geographical information systems and model-
base management. International Journal of Geographic Information Systems, 11(4), 337–357.
Beven, K. J. (1997). Distributed modelling in hydrology: applications of TOPMODEL. Chichester: Wiley.
Beven, K. J., Kirkby, M. J., Schoffield, N., & Tagg, A. (1984). Testing a physically-based flood forecasting
model (TOPMODEL) for three UK catchments. Journal of Hydrology, 69, 119–143.
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 29

Beven, K. J., Lamb, R., Quinn, P. F., Romanowicz. R., & Freer, J. (1995). Computer models of watershed
hydrology. In V. P. Singh, TOPMODEL (Chapter 18, pp. 627–668), Colorado, USA: Water Resources
Publications.
Bhargava, H., & Tettlebach, C. (1997). A web-based decision support system for waste disposal and
recycling. Computer, Environment, and Urban Systems, 22(2), 137–155.
Bundock, M. S., & Raper, J. F. (1991). GIS customisation: from tools to efficient working systems. In
Proceedings of the mapping awareness ’91 conference, London, 6–8 February (pp. 101–114).
Burrough, P. A. (1996). Opportunities and limitations of GIS-based modelling of solute transport at the
regional scale. In D. L. Corwin, & K. Loague, Applications of GIS to the modelling of non-point source
pollutants in the vadose zone (pp. 19–38). Madison, WI, USA: Special SSSA Publications.
Burrough, P. A. (1997). Environmental modelling with geographical information systems. In Z. Kemp,
Innovations in GIS 4, fourth national conference on GIS research UK (GISRUK) (pp. 143–153). London,
UK: Taylor & Francis.
Carver, S., Frysinger, S., & Reitsma, R. (1996). Environmental modelling and collaborative spatial deci-
sion-making: some thoughts and experiences arising from the I-17 meeting. In Proceedings 3rd interna-
tional conference and workshop on integrating geographical information systems and environmental
modelling. Santa Fe, New Mexico, January 1996. CD-ROM and WWW.
Clarke, M. (1990). Geographical information systems and model based analysis: towards effective deci-
sion support systems. In H. J. Scholten, & J. C. H. Stillwell, Geographical information systems for urban
and regional planning (pp. 165–175). Kluwer Academic Publishers, Dordrecht, The Netherlands.
Clarke, R. T. (1973). A review of some mathematical models used in hydrology, with observation on their
calibration and use. Journal of Hydrology, 19(1), 1–20.
Council of European communities Directive (1990). Council directive (90/313/EEC) of 7 June 1990 on the
freedom of access of information on the environment (L158; pp. 65–58, 23 July). Official Journal of the
European Communities.
Dai, Q., Evans, L., & Shank, M. (1997). Internet interactive GIS and public empowerment. In Proceedings
of integrating spatial information technologies for tomorrow, GIS ’97 (pp. 555–559). Vancouver, Canada:
GIS World, Inc.
DeBarry, P. A., Quimpo, R. G., Garbrecht, J., Evans, T. A., Garcia, L., Johnson, L. E., Jorgeson, J.,
Krysanova, V., Leavesley, G., Maidment, D. R., Nelson, E. J., Ogden, F. L., Olivera, F., Seybert, T.
A., Sloan, W. T., Burrows, D., Engman, E. T., Binger, R., Evans, B. M., & Theurer, F. (1999). GIS
modules and distributed models of the watershed (Special Report, July 1999). ASCE Task Committee on
GIS Modules and Distributed Models of the Watershed.
Deckers, F. (1993). EGIS, a geohydrological information system. In Application of geographic information
systems in hydrology and water resources (IAHS publication no. 211; pp. 611–619). Velp, Netherlands:
IAHS.
Djokic, D., Beavers, M.A., & Deshakulakarni, C. K. (1994). ARC/HEC2: an ARC/INFO–HEC-2 inter-
face. In Proc. 21th water resources planning and management division annual specialty conference (pp.
41–44). New York: ASCE.
Dodson, R. D. (1993). Advances in hydrological computation. In D. Maidment, Handbook of hydrology
(pp.23.1–23.24). New York: McGraw Hill.
Doyle, S., Dodge, M., & Smith, A. (1998). The potential of web-based mapping and virtual reality tech-
nologies for modelling urban environments. Computer, Environment and Urban Systems, 22(2),
137–155.
Feldman, A. D. (1994). Assessment of forecast technology for flood control operation. In G. Rossi,
N. Harmancioglu, & V. Yejevich, Coping with floods (pp. 445–458). Dordrecht: Kluwer Academic
Publishers.
Frank, A. U. (1993). The use of geographical information system: the user interface is the system.
In D. Medyckyj-Scott, & H. Hearnshaw, Human factors in geographical information systems (pp. 3–14).
London: Belhaven Press.
Frust, J., Gristmair, G., & Nachtnebel, H. P. (1993). Application of GIS in decision support systems for
groundwater management. In Application of geographic information systems in hydrology and water
resources (IAHS publication. No. 211). Velp, Netherlands: IAHS.
30 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

Geary, D., & McClellan, A. (1997). Graphic Java: mastering the AWT. Mountain View, California, USA:
Microsystems.
Goodchild, M. F., Parks, B. O., & Steyaert, L. T. (1993). Geographic information systems and environ-
mental modelling. Oxford: Oxford University Press.
Goonetilleke, A., Jenkins, G. A. (1999). The role of geographical information systems in urban
hydrological modelling. Journal of the Chartered Institution of Water and Environmental Management,
pp. 200–206.
Gutz, S. (1998). Up to speed with swing: user interfaces with Java foundation classes. Greenwich, CT, USA:
Manning Publications Co.
Haggett, C. (1998). An integrated approach to flood forecasting and warning in England and Wales.
Journal of Water and Environment Management, 12(6), 425–432.
HEC (1996). Status of HEC next generation software development project, US Army Corps of Engineers,
(Technical Paper No. 156; June 1996).
James, W. P., Robinson, C., & Bell, J. F. (1993). Radar-assisted real-time flood forecasting. Journal of
Water Resources Planning and Management, ASCE, 119(1), 32–44.
Jankowski, P., Nyerges, T., Smith, A., Moore, T. J., & Horvath, E. (1997). Spatial group choice: a SDSS
tool for collaborative spatial decision-making. In P. Fisher, K. C. Clark, & B. Lees, International
Journal of Geographical Information Science (Vol. 11; No. 6, September 1997).
Karimi, H. A., & Blais, J. A. R. (1997). Current and future direction in GISs. Computer, Environment and
Urban Systems, 20(2), 85–97.
Karimi, H., & Chapman, M. (1997). Real-time GISs: an emerging technology through the integration of
GPS, video imagery, and fast algorithms. In Proceeding of integrating spatial information technologies
for tomorrow, GIS ’97 (pp. 630–633). Vancouver, Canada: GIS World Inc.
Karimi, H. A., & Houston, B. (1997). Evaluating strategies for integrating environmental models with
GIS: current trends and future needs. Computer, Environment and Urban Systems, 20(6), 413–425.
Kingston, S., Carver, S., Evans, A., & Turton, I. (2000). Web-based public participation geographical
information systems: an aid to local environment decision-making. Computer, Environment and Urban
Systems, 24, 109–122.
Knill, J. (1993). Geographical information systems: the environmental view. In P. Mather, Geographical
information handling-research and applications (pp. 7–15).Chichester, UK: John Wiley & Sons.
Kuhn, W., Willauer, L., Mark, D. M., & Frank, A. U. (1992). User interfaces for GIS: discussion at the
specialist meeting (Report on the Specialist Meeting, NCGIA Research Initiative 13, 22–26 June.
NCGIA Report 92/3). National Center for Geographic Information and Analysis, University of Cali-
fornia at Santa Barbara.
Kutschera, P., Schimak, G., & Humer, H. (1996). Environmental information for public access, In R.
Denzer, D. Russell, & G. Schimak, Environmental software systems (pp. 199–208). London, UK:
Chapman & Hall.
Latkovich, V. J., & Leavesley, G. H. (1993). Automated data acquisition and transmission. In D. Maid-
ment, Handbook of hydrology (pp. 25.1–25.21). New York, USA: McGraw Hill.
Maidment, D. R. (1993). GIS and hydrological modelling. In M. F. Goodchild, B. Parks, & L. Steyaert,
Environmental modelling with GIS (pp. 147–167). New York, USA: Oxford University Press.
Maidment, D. R. (1996). Environmental modelling within GIS. In M. Goodchild, & L. Steyaert, GIS and
environmental modelling (pp. 315–323). New York, USA: Oxford University Press.
Mailhot, A., Rousseau, A., & Massicotte, S. (1997). A watershed-based system for the integrated
management of surface water quality: the GIBSI system. Water Science Technology, 36(5),
381–387.
McDonald, M. G., & Harbaugh, A. W. (1988). Amodular 3-dimensional finite difference groundwater flow
model. Techniques of water resource investigations of the USGS. Reston, CA: USGS (Book 6, Chapter
A-1\27).
Mimikou, M. A., & Baltas, E. A. (!996). Flood forecasting based on radar rainfall measurements. Journal
of Water Resources Planning and Management, ASCE, 122(3).
MMS manual (1998). The modular modelling system (MMS): user’s manual. Available at: http://
wwwbrr.cr.usgs.gov/projects/SW_precip_runoff/mms/.
 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32 31

Moore, I. D. (1996). Hydrological modelling and GIS. In M. F. Goodchild, B. O. Parks, & L. T. Steyaert,
GIS and environmental modelling: progress and research issues (pp. 143–148). Fort Collins, CO: GIS
World Books.
Moore, I. D., Turner, A. K., Wilson, J. P., Jenson, S. K., & Band, L. (1993). GIS and land surface-
subsurface process modelling. In M. F. Goochild, B. Parks, & L. Steyaert, Environmental modelling with
GIS (pp. 196–230). New York, USA: Oxford University Press.
NCGIA (1995). Collaborative spatial decision-making (Scientific Report for the Intiative-17, Specialist
Meeting).
NCGIA (1996). GIS and society: the social implications of how people, space, and environment are repre-
sented in GIS (Scientific Report for the Intiative-19 Specialist Meeting).
Nelson, E. J., Jones, N. L., & Jorgeson, J. D. (1995). A comprehensive environment for watershed
modelling and hydrological analysis. In Water resources engineering, proceedings of the first
international conference August 14–18 (pp. 829–833). San Antonio, TX: American Society of Civil
Engineers.
Nelson, E., Norman, J., Jones, L., & Woodruff Miller, A. (1994). An algorithm for precise drainage basin
delineation. ASCE Journal of Hydraulic Engineering, March, 298–312.
Oaks, S., Wong, H., & Loukides, M. (1999). Java threads. California, USA: O’Reilly & Associates, Inc.
Obled, Ch., Wendling, J., & Beven, K. (1994). The sensitivity of hydrological models to spatial rainfall
patterns: an evaluation using observed data. Journal of hydrology, 159, 305–333.
Parker, D. J., Fordham, M. H., Tunstall, S. M., & Ketteridge, A. (1995). Flood warning systems under
stress in the United Kingdom. Disaster Prevention and Management, 4(3), 32.
Parson, S. C. (1999). Development of an internet watershed educational tool (INTERWET) for the Spring
Creek watershed of central Pennsylvania. PhD thesis, in Agricultural and Biological Engineering, The
Pennsylvania State University.
Peuquet, D., Davis, J. R., & Cuddy, S., (1993), Geographic information systems and environmental
modelling. In A. J. Jakeman, M. B. Beck, & M. J. McAleer, Modelling change in environmental systems
(pp. 543–556). New York, USA: John Wiley & Sons.
Pickles, J. (1995). Ground truth: the social implications of geographical information systems. New York:
Guilford Press.
Pieplow, B. (1998). Using multi-media technology to build public consensus. Planners casebook (Vol. 25).
Chicago: American/institute of Certified Planners.
Raper, J. F., (1991), User interface for GIS. In I. Masser, & M. Blackemore, Geographic information
management: methodology and applications (pp.102–114). Harlow: Longman.
Raper, J., & Bundock, M. (1993). Development of a generic spatial language interface for GIS. In P.
Mather, Geographical information handling-research and applications. Chichester, UK: John Wiley &
Sons.
Raper, J., & Livingstone, D. (1995). Development of a geomorphological spatial model using object-
oriented design. International Journal of Geographical Information Systems, 9, 359–385.
Romanowicz, R., Beven, K., & Moore, R. (1993). GIS and distributed hydrological models. In P. Mather,
Geographical information handling-research and applications (pp. 197–205). Chichester, UK: John Wiley
& Sons.
Schultz, G. A. (1993). Application of GIS and remote Sensing in hydrology and water resources, Vienna,
1993. IAHS Publication, 211, 127–140.
Stewart, K. G. (1999). Managing and distributing real-time and archived hydrological data for the Urban
Drainage and Flood Control District’s ALERT system, 26th annual water resources planning and
management conference, American Society of Civil Engineers, June 6–9, 1999, Tempe, Arizona.
Stuart, N., & Stocks, C. (1993). Hydrological modelling within GIS: an integrated approach. Application
of geographic information systems in hydrology and water resources (IAHS publication no. 211; pp. 319–
329). Velp, Netherlands: IAHS.
Sui, D. Z., & Maggio, R. C. (1999). Integrating GIS with hydrological modelling: practices, problems, and
prospects. Computers, Environment and Urban Systems, 23, 33–51.
Taylor, K., Cameron, M., & Haines, J. (1998). An integrated information system on the web for watershed
management. ACM GIS’98, November 1998, Washington, DC, USA.
32 W. Al-Sabhan et al. / Comput., Environ. and Urban Systems 27 (2003) 9–32

Taylor, K., Ackland, R., Walker, G., & Jackson, D. (1997). WWW delivery of spatio-teporal modelling
result for watershed management: a case study. In A. D. McDonald, & M. McAleer, MODSIM 97,
International congress on Modelling & simulation proceedings (pp. 1574–1579, 8–11 December, 1997).
Thoen, B. (1995). Web GIS: toy vs. tool. Available at: http://www.gisnet.com/gis/Webgis.html.
Van Deursen, W. P. A. (1995). Geographical information systems and dynamic models, development and
application of a prototype spatial modelling language. Dissertation, Utrecht, Netherlands.
Walker, S. (1991). Keys to successful implementation of computer aided decision support systems in water
resources management. Advances in Water Resources Technology, 467–472.
Wesseling, C., Karssenberg, D., Burrough, P., & Van Deursen, W. (1996). Integrating dynamic environ-
mental models in GIS: the development of a dynamic modelling language. Transactions in GIS, 1(1),
40–48.
Wilson, J. P. (1996). GIS-base and land surface/subsurface modelling: new potential for new model? Third
international conference on GIS and environmental modelling, Santa Fe, NM, USA.
WMS reference (1999). WMS watershed modelling system, reference manual. Brigham Young University-
Environmental Modelling, Research Laboratory.
Wood, D., Gooch, T. C., Pronovost, P. M., & Noonan, D. C. (1985). Development of a flood manage-
ment plan. Journal of Water Resources Planning and Management, 111(4), October 1985, pp. 417–433.
Worboys, M. F. (1995). GIS: a computing perspective. London: Taylor & Francis.

Further reading

Burrough, P. A. (1986). Principles of geographic information systems for land resources assessment. Oxford:
Clarendon Press.
Goodchild, M. F., Haining, R., & Wise, S. (1992). Integrating GIS and spatial data analysis: problems
and possibilities. International Journal of Geographic Information Systems, 6, 407–423.
Gosling, J., & McGilton, H. (1996). Java language environment: a white paper. Available at: http://
java.sun.com/docs/white/langenv. Sun Microsystems.
HEC (1998). HEC-HMS hydrological modelling system, user’s manual. Davis, CA: US Army Corps of
Engineers.