Sajn, Robert and Stafilov, Trajče and Balabanova, Biljana and Alijagic, Jasminka (2021) The assessment of spatial distribution of trace elements in soil and moss using artificial intelligence in the Bregalnica river basin, North Macedonia. Mining and environmental protection proceedings. ISSN 978-86-7352-372-9
Text
PROCEEDINGS_MEP21_FINAL-Sajn-Stafilov-Balabanova-2021.pdf Download (1MB) |
Abstract
The main objectives of this paper are to improve spatial distribution maps using an artificial neural
network - multilayer perceptron (ANN-MLP) at macroregional scale and to compare the two methods, the universal
kriging method and ANN-MLP for representing the spatial distribution of selected anthropogenic and geogenic
elements in North Macedonia. The Bregalnica River Catchment Area (5400 km2) is heavily contaminated due to
long-lasting mining activities (Figure 1). There are many agricultural lands in the region that are irrigated with the
contaminated Bregalnica River that flows past the mines. The level of contamination in two different sampling
media, soil and moss samples have been measured by using ICP-MS and ICP-ES. The soil and moss measurements
together with geospatial data obtained from the digital elevation model (DEM), land use data, and remote sensing
are incorporated into the spatial distribution mapping using the advanced prediction modelling technique ANNMLP.
Both methods have been further compared and evaluated, which lead to the general conclusion that ANNMLP
gives more realistic and detailed results than the universal kriging method used for the reconstruction of main
distribution pathways.
Item Type: | Article |
---|---|
Subjects: | Natural sciences > Chemical sciences Natural sciences > Earth and related environmental sciences Agricultural Sciences > Other agricultural sciences |
Divisions: | Faculty of Agriculture |
Depositing User: | Biljana Balabanova |
Date Deposited: | 22 Nov 2021 12:20 |
Last Modified: | 22 Nov 2021 12:20 |
URI: | https://eprints.ugd.edu.mk/id/eprint/28798 |
Actions (login required)
View Item |