Element enrichment factor calculation using grain-size distribution and functional data regression.
Chemosphere
; 119: 1192-1199, 2015 Jan.
Article
em En
| MEDLINE
| ID: mdl-25460761
In environmental geochemistry studies it is common practice to normalize element concentrations in order to remove the effect of grain size. Linear regression with respect to a particular grain size or conservative element is a widely used method of normalization. In this paper, the utility of functional linear regression, in which the grain-size curve is the independent variable and the concentration of pollutant the dependent variable, is analyzed and applied to detrital sediment. After implementing functional linear regression and classical linear regression models to normalize and calculate enrichment factors, we concluded that the former regression technique has some advantages over the latter. First, functional linear regression directly considers the grain-size distribution of the samples as the explanatory variable. Second, as the regression coefficients are not constant values but functions depending on the grain size, it is easier to comprehend the relationship between grain size and pollutant concentration. Third, regularization can be introduced into the model in order to establish equilibrium between reliability of the data and smoothness of the solutions.
Palavras-chave
Texto completo:
1
Coleções:
01-internacional
Base de dados:
MEDLINE
Assunto principal:
Poluentes do Solo
/
Modelos Lineares
/
Química
/
Interpretação Estatística de Dados
/
Ecologia
/
Poluição Ambiental
/
Modelos Teóricos
Tipo de estudo:
Prognostic_studies
Aspecto:
Determinantes_sociais_saude
País/Região como assunto:
Europa
Idioma:
En
Revista:
Chemosphere
Ano de publicação:
2015
Tipo de documento:
Article
País de afiliação:
Equador
País de publicação:
Reino Unido