Multivariate statistical-and gis-based approach developed for integrated environmental analysis in urban watershed

• Autor(a) principal: Urban, Rodrigo Custódio
• Data de Publicação: 2018
• Outros Autores: Manfré, Luiz Augusto, da Silva, Alexandre Marco [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• DOI: 10.4090/juee.2018.v12n1.077087
• Texto Completo: http://dx.doi.org/10.4090/juee.2018.v12n1.077087; http://hdl.handle.net/11449/221242
• topic: Environmental planning; Environmental zoning; Multiple correspondence analysis; Principal component analysis; Urban soil quality; Urban water quality

This paper aims to assess the environmental quality of a small urban watershed, located at a sub-tropical region highly urbanized in Brazil, using water and soil quality, land cover and terrain characteristics. The proposed methodology was based in physical and chemical features of 40 soil sampling sites, land cover and slope. Principal Components Analysis (PCA) was used to define the best variables to the analysis. The soil quality, land cover and slope data was grouped and categorized in qualitative variables. Multiple Correspondence Analysis (MCA) was applied to cluster the variables. Geographic Information System (GIS) tools were used to build the zoning map. During 12 months water was sampled in two sites in the same river at the watershed. PCA was used again to define water quality and differences between the two sampling sites. Porosity and carbon rate were the principal soil variables to distinguish three different soil zones 1, 2 and 3 represent 15,1%; 9,8% e 75,1% of the area, respectively. Zone 1 present condition that must be conserved to maintain environmental services as water retaining and carbon storage. Related to water quality, the PCA presented differences between dry and wet season. Besides, sampling site 1, located within a vegetation region presents better conditions than sampling site 2, located within urban land cover. The assessment method used multivariate statistics and GIS. The methodology is a useful tool to environmental planning. The replication of this methodology is encouraged, in order to assess its suitability in different conditions, i.e. climate and size.