![]() |
|||
![]()
|
![]() |
![]() |
![]() |
5.5. COLOSSDoneker et al. Columbia River Decision Support System (COLOSS) is a simple, but representative expert advisor currently under development for U.S. Army Corps of Engineers. Its objective is to provide support for decision-makers that anticipate a competing set of use for the water resources of the Columbia River. The system Like CORMIX it is being built by Doneker (Doneker 96) using production rules as the dominant representation schema. COLOSS will provide decision support for management of the competing water resource requirements of the lower Columbia River system. The rule base will include the physical and regulatory constraints for river and reservoir operation. COLOSS will aid in reservoir operation for water supply, navigation, hydropower production, recreation, flood control, and fishery needs. The intent of the project is to demonstrate the capability of a rule-based system for decision support of river operations. In general, the knowledge base that COLOSS will operate is to include aspects of use, timing, depositions, mixing, and other elements related to the varied use of the water resources in the area. The software will be linked to the Columbia River Operation Hydromet Management System (CROHMS). Results will be conveyed via tables, messages, and graphics on a window-based system. 6. ECOSYSTEM MANAGEMENT6.1. INTRODUCTIONEcosystem management clinches nearly every aspect of natural resource management. Even topics addressed by the systems within the mineral, water, and forestry domain are encompassed by this type of administration. Its scope is tremendous, which is well illustrated by Short et al. (Short 95) in their report on Earth System Science (ESS). Although ecosystem management may impose a regional locus of interest, it is becoming more evident that its character is truly global. Rainforest management in Brazil can no longer be considered a national issue as long as these forests influence the global climate. When the Tchernobyl disaster in 1986 threatens the wildlife habitat of the Laps in the northern regions 10 years later, nuclear power for regional supply is no longer an issue only for Russians. Moreover, if the herring in the North Atlantic are over-taxed by Spanish, Portuguese, Scottish, Norwegian, and Russian trawlers, it is not solely a case for the EU fishing interests. It will ruin the balance in the ocean, threatening for instance the subsistence of cod, seal, and birds, and the livelihood of coastal communities in Greenland, Iceland, and northern Norway. True ecosystem management must be supernational and keep a sustained focus on nature's genuine diversity. Fundamental to all kinds of modern ecosystem management are monitoring and control. This calls for a tremendous amount of data about almost every aspect of our planet. In a more limited sense, it addresses monitoring and control of pollution. Environmental impact assessment (EIA) and environmental protection are both essential. The former deals with how our resources are maintained and exploited. Environmental protection issues embrace both pollution restriction, contamination from hazardous materials, and habitat shielding. Due to its increased importance and expanding complexity and scope, there is an increased need for more tools to aid the management task. Expert systems that assist in determining the effects on the environment due to various industrial, agricultural, or municipal initiatives have arisen. The work of Hushon (Hushon 91) is an early example of that. The system made by Sturrock et al. (Sturrock 94) shares some of the same intention, but sees the issue from the point of the industry. In other words, mismanagement of production plants may lead to pollution and thus create problems for the industry. That type of system is not discussed here. "Think globally, act locally" is a slogan that can be applied in conjunction with ecosystem management. In undeveloped countries, fire is often used as a farming tool to clear new areas and fertilize new crops. Applied on a large scale, this practice can have devastating effects on both farmland and the surrounding wildlife. One major implication is soil erosion and dehydration of the soil, which paves the way for exhaustion and desert-like conditions. Although harmful to a large area, management practices must be distributed to local areas. Knowledge sharing is a key word in this context, and the early work on a system for vegetation succession advice by Dr. Noble (Jeffers 85) stands out as a very good example. Global or large area monitoring today implies some kind of remote sensing. Some of the expert systems in the geological domain have been built to support remote sensing for mineral exploration. We have also seen an example of this in forestry. Its application is general and paves the way for a complete overview of the whole earth and multiple aspects of an ecological system. The work undertaken by NASA (Short 95) is clearly the most ambitious of all the initiatives reported in this direction. Remote sensing inevitably calls for advanced image processing. Every expert system dealing with this addresses the image processing task to create meaningful representations. However, in terms of ecosystem management, intelligent data fusion is perhaps an equally important issue. The nature of the management task discussed here is complex and calls for many types of support. Hence, most systems within this genre are hybrid systems applying a wide variety of computing resources. GIS is central, advanced database technology; simulation in addition to a diverse set of knowledge bases are common elements. Neural networks have moved into application (Wong 93; Boddy 94). Web technology and multimedia often support the delivery of expertise. An extreme version of a hybrid approach is documented in the description of the underwater robot system documented by Stoker et al. (Stoker 95). Here, artificial intelligence supports the use of virtual reality and telepresence technology to explore and monitor an underwater habitat in Antarctica.
In Table 4 we have included a list of systems that address ecosystem management at different levels. The 10 years between Noble's system and the IIFS represents a leap in technology. However, both systems serve a good purpose within their own scope. The comparison acts to emphasize that expert system technology must be suited for its use -- namely, to deliver expertise in a situation or place where it is otherwise not immediately available.
|
![]() |
|
Use of this site is subject certain Terms & Conditions. Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details. |