genetic algorithm to reduce stream channel cross section data

• Published in: Journal of the American Water Resources Association Vol. 42; no. 2; pp. 387 - 394
• Main Author: Berenbrock, C
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01.04.2006; American Water Resources Association
• Subjects: algorithms; case studies; computational methods; cross section; data analysis; data management; data reduction; Earth sciences; Earth, ocean, space; equations; Exact sciences and technology; genetic algorithm; Hydrology; Hydrology. Hydrogeology; rivers; simulation; streams; surface water; Water resources; Montana; British Columbia; Idaho; Canada; United States; Western Canada; rivers; computational methods; algorithms; models; data management; software; cross section; simulation; streams; digital simulation; North America; surface water; genetic algorithm; data reduction; water resources; channels; water resource management; programs; methodology
• ISSN: 1093-474X; 1752-1688
• DOI: 10.1111/j.1752-1688.2006.tb03845.x

A genetic algorithm (GA) was used to reduce cross section data for a hypothetical example consisting of 41 data points and for 10 cross sections on the Kootenai River. The number of data points for the Kootenai River cross sections ranged from about 500 to more than 2,500. The GA was applied to reduce the number of data points to a manageable dataset because most models and other software require fewer than 100 data points for management, manipulation, and analysis. Results indicated that the program successfully reduced the data. Fitness values from the genetic algorithm were lower (better) than those in a previous study that used standard procedures of reducing the cross section data. On average, fitnesses were 29 percent lower, and several were about 50 percent lower. Results also showed that cross sections produced by the genetic algorithm were representative of the original section and that near-optimal results could be obtained in a single run, even for large problems. Other data also can be reduced in a method similar to that for cross section data.