In the context of this review the DSSs will be analysed in their ability to support the SEA in its objective of developing an overall framework approach to ensure South Africa’s water resources are utilised optimally in the short and long term to the best benefit of the people and environment of South Africa.

In this project the definition of a DSS will be formulated in terms of a system that will provide the information required to service water managers at the CMA level in terms of the NWA (1998). It must also be stated that in the course of reviewing the different DSSs currently developed or in place, that it is unlikely that they will fulfil the requirements of the SEA and CMAs as they will have been developed to address different problems and concerns.

The review in this document is based on the review given in a WRC report by Görgens and Jewitt (2000) which was based on the review and testing of the different DSSs which were available in 1995 when the tests took place. Where possible new DSSs have been included in the review and updated information has been obtained on some of the systems. In this section common features pertaining to the various different DSSs are included and their application in terms of the requirements of CMAs will be assessed. Short reviews of all the different DSSs follows.

Catchment Management Support System (CMSS) was developed in response to the needs of various organisations involved in catchment management in Australia (CSIRO, 1994, as cited in Jewitt and Görgens, 2000). It is a computer program, operating on a PC workstation in a graphical environment, used to assess likely changes in nutrient loads entering streams as a result of catchment landuse changes. Its design philosophy is based on the implementation of simple relationships between selected components. Input data are landuse, nutrient generation rates and management practice data. Output is a simple function of these inputs and is presented in the form of high quality colour graphs and maps, although the system does not place great emphasis on the GIS link.

The lack of linking to a GIS and the inability to incorporate other variables such as economic and social criteria are limiting factors for the use of this system.

The aim of this system is the automation of information retrieval in support of complex GIS query, analysis and modelling. This system has the potential for the analysis of the spatial distribution of temporal trends in climate events, modelling spatio-temporal interactions in meteorological events and assessing the socio-economic impacts of extreme events in time and space depending on the information available (Yan, 2001).

This system is still in the developmental stages, although it does offer many of the advantages of linking GIS with modelling and could perhaps be an appropriate methodology to adopt in the design of the WMA DSS.

CRAM was developed as an integrated catchment management system by the CSIR using the Crocodile River as a prototype (Meyer and Scholes, 1994, as cited in Jewitt and Görgens, 2000). It is a hydrologically based system, which allows for the simulation of environmental, social, economic and hydrological impacts resulting from changes in the catchment.

The system is designed to be easy to use and navigate, with user input being guided by Graphical User Interfaces (GUIs). The system is linked to GIS and utilises an efficient database management system.

The major weakness of the model is the simplistic algorithms used to simulate the hydrology, in particular, but also in the other aspects that are been simulated (Jewitt and Görgens, 2000). CRAM could be used as a simple scenario generation or gaming tool, however, the model output cannot necessarily be accepted as accurate. The CRAM system is highly simplified and has become outdated.

• G2-AEAM

The G2 model is developed by utilisation of the adaptive environmental assessment and management process (Holling, 1976; Walters, 1986 as cited in Jewitt and Görgens, 2000). The AEAM process involves the development of simulation models in a multi-disciplinary workshop environment. Modelling algorithms are developed for predefined spatial and temporal scales by subgroups of the workshop. The algorithms that have been developed by the different subgroups are then returned to the workshop and entered into the G2 models basic shell. The model is then run to return simple graphics operating on real time, which represent the output of the model.

The obvious weakness is the limitation of the algorithms produced and the generalisation that may occur as the model is normally run at a regional scale with monthly time steps. The methodology does, however, produce some success through the workshop environment as consensus is achieved through the various expert groupings regarding the algorithms used in the model. The model is put together easily and results are available for discussion in a simple and usable form within the workshop. Once the model has been completed and run the discussion between the different participants is reported to be very useful.

The simplified nature of the modelling makes this system unusable for the decision making needs of a CMA.

• HSPF – ANNIE

This is more of a useful modelling system than a DSS, which manages and displays data stored in the WDM. ANNIE has a text based user interface. This system has, however, become outdated and has been superseded by the HSPF – BASINS link described next.

The BASINS system was developed by the US EPA to help address water quality issues in the USA. The system uses ArcView as the framework to provide users with a fully comprehensive catchment management tool to particularly address water quality issues. BASINS was developed to take advantage of recent developments in software, data management technologies and computer capabilities. BASINS addresses three primary objectives which are

BASINS consists of interrelated components required for catchment and water quality analysis such as local databases, assessment tools, models and post-processing output tools. The system has strong links to the HSPF model. The system is being constantly updated and new versions are released from time to time.

The BASINS-HSPF link offers many opportunities on the catchment management front. There are, however, limitations in terms of the database management (cf. Section 7.6). The system does not include socio-economic data and as it is not locally developed the incorporation of these variables could be both time and resource consuming.

• HYMAS

HYMAS (Hughes, 1994, as cited in Jewitt and Görgens, 2000) is a "DOS based shell used to operate several hydrological related models" which include the VTI and PITMAN models. HYMAS has a menu-driven user interface, which is both user friendly and easy to use, although at this stage it is only keyboard driven. This model uses a binary database, which has the ability to store data in a variety of a time steps, and allows the user to manipulate time series and edit input information for a variety of different models. HYMAS has a number of post-processing options, which allow for numerous different useful output options to be used.

HYMAS is restricted to the DOS platform and has no GIS link. It is, however, well written, easy to use and offers sensible options to the user. It does not at present incorporate social and economic data.

• MIKE – 11

MIKE-11 is a professional engineering software package, which incorporates modules for the simulation of flows, water quality and sediment transport in rivers, estuaries, irrigation systems, channels and other water bodies. MIKE-11 offers features such as a fast and robust numerical scheme; advanced cohesive and non-cohesive sediment transport modules as well as comprehensive water quality and eutrophication modules (MIKE-11, 2001). In addition, this model provides links to advanced hydrological, sewer and coastal modelling tools. A simple SCS based rainfall-runoff module is also included in the package. This system can be used under both DOS and Windows and is relatively user-friendly. Physical catchment and channel parameters are input into the model and the system allows the import and export of data in a number of commonly used ASCII formats. Graphical representation of simulated results is possible within this system.

Time series management is, however, not available and no land based water quality simulations are possible. A major disadvantage of MIKE-11 is its high cost.

• The Modular Modelling System

The Modular Modelling System (MMS) (Leavesley et al., 1994, as cited in Jewitt and Görgens, 2000) "has been developed as a common framework in which to focus multi-disciplinary work". MMS is operated through an X-windows based GUI on a UNIX based workstation. Data may be geographically represented and manipulated by means of the pre and post processing tools available. The system provides tools for linking process modules representing various catchment processes. A GIS link has also been developed. Other modules may be added provided they conform to the programming norms. Such modules may be written in FORTRN or C.

The model however offers little user-support and is somewhat complex requiring an extensive amount of time to become proficient in the system.

• RAISON

RAISON (Regional Analysis by Intelligent Systems ON microcomputers) is a Windows software package developed for ecosystem management on a catchment basis. It is a versatile environmental information system with built in expert system capabilities designed specially for decision support applications. RAISON offers a broad framework that allows for the integration of data, spreadsheets, text, maps, objects and knowledge input. In addition, there are also numerous user-friendly tools offered by the system thus enabling the user to produce output for interpretation, integration, advice, classification, analysis and recommendations. It is particularly suitable for applications which involve point (e.g. monitoring sites) and polygon (e.g. land use) data.

RAISON is able to import from, or export to, many commercially available databases, spreadsheets and GISs. RAISON includes an expert system for knowledge acquisition to provide a link between numeric and descriptive knowledge which is required for decision support and environmental information. It also offers a map-based GUI, which is customised for each application.

The RAISON software package is the most comprehensive DSS tool analysed in this document. It offers the integration of economic, environmental, social and hydrological data. This software is however relatively expensive and there is no local support available.

• SWAMP – HYDRA

The HYDRA system (Davies et al., 1994, as cited in Jewitt and Görgens, 2000) has been developed in order to connect pre-existing modules. The proto-type system SWAMP has been developed around the HSPF model. The HYDRA system builds upon advances, particularly in the in the field of databases, in the integration of GIS with other models. The system comprises a GUI developed on the basis that the water resources manager is the primary user and as a result options revolve around likely management questions. The central component of HYDRA is a system manager, which controls all model-model and model-user communications. The systems manager has its own local database which stores the data required by the various active model components. A library of existing models, each with its own driver, is also a component of HYDRA, each of which may be integrated by the system.

A disadvantage of this seemingly sensible system is the high level of computer science and information technology required.

• WDM Guide

The WDM Guide was developed as a prototype information system (Van Rensburg and Dent, 1997). The software runs under Windows or on a UNIX workstation and is based on ArcView Version 2 and its associated Avenue programming language. It provides flexibility to the user by utilising the multi-tasking facilities of the UNIX and Windows operating system. The system is user-friendly; it has a mouse and menu driven interaction with time series stored centrally in a WDM file on a workstation accessible through the Internet as well as image and ARC/INFO data sets, which reside locally on a PC.

Features of the WDM Guide include graphical query of selected variables at any point in the selected stream network. It is also able to query and plot a number of variables simultaneously. The system allows the query and display from available images per sub-catchment as well as the animation of monthly time-series of a selected variable. In addition data is accessible via remote networking.

The WDM Guide is based largely on easily available commercial software, to which simple functionality may be added by use of public domain software tools. This software has been outdated and superseded by the use of the BASINS-HSPF system.

Out of all the systems analysed possibly the most promising options are provided by the RAISON modelling system and the Mike 11. Both systems are able to cope with data input and query and store data in a spatially referenced format. The Mike 11 model seems to concentrate more on the hydrological side of data capture and manipulation with a suite of models accessing the database. The RAISONs system looks like it would be able to cope with many different data formats and could possible be used to store and manipulate social and economic data. As it has been described as been able to store and manipulate both numeric and descriptive data. While no system would be ideal in solving the problems introduced by the NWA (1998) the RAISON system looks the most promising. The system is however extremely expensive which could make its use prohibitive and there is very little local support available. The best options in terms of the DSS is perhaps to develop the system from scratch making sure that it has the ability to cope with economic, hydrological, social and environmental data. The modelling tools associated to the system would then be local and designed to suit local conditions and data.