« Back
Integration of artificial neural network and field portable metals analyzers (FP-XRF) in forecasting organic micropollutants on agricultural soils
Poster Title: Integration of artificial neural network and field portable metals analyzers (FP-XRF) in forecasting organic micropollutants on agricultural soils
Submitted on 04 Apr 2017
Author(s): Maria Grazia Bonelli and Andrea Manni
Affiliations: University of Rome La Sapienza, Chemical Research 2000 S.r.l.
This poster was presented at PITTCON 2017
Poster Views: 1,171
View poster »

Poster Information
Abstract: The assessment of organic and inorganic pollution in agricultural soils is a difficult challenge due to the large area used to collect and analyse a lot of samples. The measurement of inorganic compound, such as heavy metals, can be obtained by analytical techniques for on-site screening and fast analysis of metals in soils and sediments like Field portable X-ray fluorescence (FPXRF) spectrometry. This method is no destructive and allows a rapid qualitative pollution characterization, allowing to identify the hot spots where to extract the soil samples to be analysed in the laboratory by traditional techniques. However, a fast and low-cost method to measure organic compounds, such as dioxins and PCBs, in alternative to laboratory techniques actually doesn’t exist. The aim of the present study has been to find a method able to screening the pollution of an agricultural land, using a Field Portable X-ray fluorescence to measure metals value and a statistical procedure, such as the Artificial Neural Networks technique, to estimate unknown concentrations of dioxins and PCBs trough the relationships between the organic and inorganic pollutants.
Summary: ANN algorithm is an efficient method for forecasting POPs concentration from inorganic compounds due its features which is an advantage to estimate variables also in absence of linearity. The use of ANN coupled to a metal portable soil analyser, such as FPXRF could be a valid technique to screen large polluted agricultural soils, identifying the hot spots more quickly and cheap than with traditional and geo-statistical techniques.
References: Azadi S., Karimi-Jashni A. (2016) in Wast management 48, 14-23
Demuth H., Beale M., Hagan M. (2008) Neural Network ToolboxTM6. User’s guide MathWorks 9 (4), 259-265
ERSAF-Regione Lombardia, “Analisi del contenuto in rame ed altri metalli nei suoli agricoli lombardi”, in Quaderni della ricerca 61, 23-51 (2007)
Fiedler H. Dioxins and Furans (PCDD/PCDF) in The Handbook of Environmental Chemistry Vol. 3, Part O Persistent Organic Pollutants - (2003) © Springer-Verlag Berlin Heidelberg
H. Wackernagel, Multivariate Geostatistics: an introduction with applications, Springer, Berlin, 161-164 (2003)
IBM SPSS, Neural networks 22.0 (2013),
Kriesel D. (2005) A Brief Introduction to Neural Networks
USEPA UPerez-Vazquez, F.J., Flores-Ramirez, R., Ochoa-Martinez, A.C., Orta-Garcia, S.T., Hernandez-Castro, B., Carrizalez-Yañez, L., and Pérez-Ma
Report abuse »
Ask the author a question about this poster.
Ask a Question »

Creative Commons