Determinação da rede de drenagem em grandes terrenos armazenados em memória externa

• Ano de defesa: 2013
• Autor(a) principal: Gomes, Thiago Luange
• Orientador(a): Andrade, Marcus Vinícius Alvim
• Banca de defesa: Gonçalves, Luciana Brugiolo , Carneiro, Tiago Garcia de Senna
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Viçosa
• Programa de Pós-Graduação: Mestrado em Ciência da Computação
• Departamento: Metodologias e técnicas da Computação; Sistemas de Computação
• País: BR
• Palavras-chave em Português: Modelagem de terrenos; SIG; Memória externa; Hidrografia
• Palavras-chave em Inglês: Land modeling; GIS; External memory; Hydrography
• Área do conhecimento CNPq: CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://locus.ufv.br/handle/123456789/2649

This work presents a very efficient algorithm, named EMFlow, and its implementa- tion to compute the drainage network on huge terrains stored in external memory. The drainage network of a terrain delineates the path that water flows through the terrain (the flow direction) and the amount of water that flows into each terrain cell (the flow accumulation). As it is known, due to the fast increase in the volume of high resolution terrestrial data available, the internal memory algorithms do not run well for huge terrains on most computers and, thus, optimizing the massive data processing algorithm simultaneously for data movement and computation has been a challenge for GIS (Geographic Information System). In EMFlow, the flow direc- tion is computed using an adaptation of method RWFlood which uses a flooding process to obtain this direction and the flow accumulation is computed based on a very fast method proposed by Haverkort and Janssen (2012). To reduce the total number of I/O operations, EMFlow adopts a new strategy to subdivide the terrains into islands which are processed separately and the terrain cells are grouped into blocks, which are stored in a special data structure managed as a cache memory. The EMFlow execution time was compared against the two most recent and most efficient published methods: TerraFlow and r.watershed.seg and it was, in average, 27 times faster than both methods. Since processing large datasets can take hours, this improvement is very significant.