Bertens, Jurjen 
  
environment and are consistent with data availability and our ability to mathematically represent hydrological systems 
(Grayson et al., 1993). 
A second problem concerns the interpretation of predictions with respect to EIA. At present the possibility for individuals 
outside the EIA team (e.g. decision makers) to run simulations of the expected impacts and evaluate consequences from their 
own perspective is rather limited (Beinat ef al., 1999). When decision makers are not familiar with the nature of the impact 
considered, they will encounter difficulties taking the predictions into account when making a decision (for example, 
someone with limited knowledge of soil erosion processes will have difficulties interpreting a value for increase in annual 
soil erosion). The need for expertise (amongst other factors, such as ideological blindness or malice) can limit a user's 
ability to retrieve the correct information from a map (Van Herwijnen, 1999). User-friendly graphics assist individuals with 
little experience in hydrological matters in running a hydrological model and producing good-looking graphics. 
Sophisticated visualisations and data handling tend to seduce the user into an unrealistic sense of model accuracy (Grayson 
et al., 1993). A potential danger is thus that digital data always appear to be of high quality and information on data quality 
and errors is either neglected or in some cases not available (Thieken et al, 1999). *Maps provide an excellent 
communications medium for presenting results in a form that most people think they can understand" (Openshaw, 1991). 
Additionally, interpretation-related problems can arise when a number of fundamentally different impact predictions have to 
be compared to each other (e.g. balancing a change in project costs with a predicted change in annual soil erosion is not that 
obvious). 
4 METHODOLOGY 
4.1 Introduction 
The development of a workable methodology for the adaptation of environmental models to their use in EIA and converting 
the model into an interactive tool that generates useful information regarding the implications of the proposed construction 
could be a very useful aid to environmental management (useful in the sense that the results can be: (1) interpreted by 
decision makers; and (2) compared to other impacts). “Information technology, and in particular the integration of database 
management systems, GIS, remote sensing and image processing, simulation and multi-criteria optimisation models, expert 
systems and computer graphics provide some of the tools for effective decision support in natural resources management.” 
(Fedra, 1995). The combination of a distributed hydrological model and the mapping capabilities of a GIS greatly reduce 
processing time for data preparation and presentation. This combination is sometimes referred to as a decision support 
system (Grayson ef al., 1993). The main rationale for the development and use of decision support systems is its power to 
reduce redundancy by summarising, categorising and projecting relevant data (Barr & Sharda, 1997). This should ideally 
decrease the amount of cognitive effort required for processing large amounts of information. 
Bathurst & O’Connell (1992) give a two-stage procedure for the application of a hydrological model within the context of a 
decision support system. In the first stage a model is set up for the required watershed and conditions. In this stage 
hydrological expertise is fundamental. In the second stage the model is applied to the evaluation of impacts of proposed 
changes. Less technical expertise will be required, since the model has been validated and implementation should be backed 
up by user-friendly support. The policy maker should then be able to examine the effect of the proposed change on the 
output attribute of interest. The representation of attributes should allow for trade-offs between environmental and socio- 
economic qualities. 
4.1 Model Development 
An approach is presented for the development of hydrological models for specific use within an EIA framework. The 
approach is divided into three main steps: (A) the development of a dynamic, spatially distributed hydrological model for a 
specific part of the hydrological regime (related to concerns) in a given area; (B) implementation a proposed project in the 
model; and (C) implementation of the model in the EIA framework for aid in decision support. These three phases are 
interrelated. That is, the models structure will depend on both nature of the infrastructure considered and desired output (if 
flood hazard is a major concern, relevant output indicators would be area potentially affected, flood frequency/level etc.). 
Changes in the hydrological regime can then be used to assess the expected effects on geomorphological processes such as 
channel flow, soil and channel erosion, waterlogging, or sediment production and transport. 
  
178 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 
  
€» tn « 
pe