Determination of field-based sorption isotherms for Cd, Cu, Pb and Zn in Dutch soils
Bepaling van op veldgegevens gebaseerde sorptie-isothermen voor Cd, Cu, Pb en Zn in Nederlandse bodems
26 May 2012, PDF |
99 pages |
Otte JG, van Grinsven JJM, Peijnenburg WJGM, Tiktak A
RIVM Report 711401007
Sorption isotherms for metals in soil obtained in the laboratory generally underpredict the observed metal content in the solid phase in the field. Isotherms based on in-situ data are therefore required. The aim of this study is to obtain field-based sorption isotherms for Cd, Cu, Pb and Zn as input for the heavy-metal accumulation model SOACAS. Two types of sorption isotherms were fitted using field data sets in this research: the Freundlich type, with one solid-phase fraction was fitted by Stepwise Linear Regression (LR), and a model with two solid phase mass fractions, a reactive fraction and a non-reactive fraction, was fitted with Nonlinear Least Squares Regression (NLSSR). From the results of the LR and the NLLSR fits of the Hoop-Janssen data set, the generally explained variance for the Zn models is shown to be the highest, followed by the Cd models. The explained variance of the Cu and Pb models is lower. The performance of the LR models and the NLLSR models is comparable. The NLLSR fits of the 2-phase isotherms of the Hoop-Janssen data set almost never include a statistically significant inert metal fraction. This implies that a 2-phase model can, in most cases, not be derived from this data set. The LR models derived from field data (this research) predicts the observed metal content in the solid phase in the field best compared with isotherms derived from batch data. The extension of the field partition data set and quality improvement is recommended for future research.