Machine Learning Models for Predicting the Ammonium Concentration in Alluvial Groundwaters
- PDF / 9,779,566 Bytes
- 17 Pages / 595.276 x 790.866 pts Page_size
- 82 Downloads / 129 Views
Machine Learning Models for Predicting the Ammonium Concentration in Alluvial Groundwaters Marija Perović 1
&
Ivana Šenk 2
&
Laslo Tarjan 2
&
Vesna Obradović 1 & Milan Dimkić 1
Received: 26 November 2019 / Accepted: 27 August 2020 # Springer Nature Switzerland AG 2020
Abstract Considering the great importance of groundwater quality for water supply, in the last decade, significant scientific attention has been devoted to nitrate reduction transformation pathways and nitrogen conservation in groundwaters in the form of ammonium. To evaluate and assess the ability of machine learning models to predict the ammonium concentration, four machine learning models were applied: a three-layer neural network (NN), a deep neural network (DNN), and two variants of support vector regression (SVR) models: with linear and with Gaussian radial basis function kernel. A dataset of 322 samples with 13 predictor variables representing selected parameters responsible for oxidative/reductive nitrogen transformations in shallow alluvial groundwater was acquired from measurements in 55 monitoring wells during a 6-year monitoring period (2011–2016) in Serbia. Applied principal component analysis and cluster analysis gave an insight into conditionality and relations between the selected parameters, distinguishing four main factors, which explained 70.97% of total variance, and classifying examined objects by similarity. Extracted factors correlated the concentration patterns, implying the main nitrogen transformations in examined groundwater. The machine learning models were successfully applied for predicting the ammonium concentration with high determination coefficients (R2) in tests: 0.84 for DNN and 0.64 for NN, while the SVR did not prove to be adequate with the best R2 of 0.24. Keywords Ammonium . Groundwater . Factor analysis . Machine learning . Neural networks
1 Introduction Since the population growth in many areas worldwide was followed by excessive application of nitrogen fertilizers and usually by sanitation network deficiency, groundwater * Marija Perović [email protected] Ivana Šenk [email protected] Laslo Tarjan [email protected] Vesna Obradović [email protected] Milan Dimkić [email protected] 1
Jaroslav Černi Water Institute, Jaroslava Černog 80, Pinosava, Belgrade 11226, Serbia
2
Faculty of Technical Sciences, University of Novi Sad, Trg Dositeja Obradovića 6, Novi Sad 21000, Serbia
pollution by nitrogen compounds has become considerable on a global scale [1]. Constant increase of nitrogen loading pressure on aquifers and hydraulically connected surface water bodies overcomes the capacity of the natural ecosystem service: water self-purification potential. Knowing the source and final compound of nitrogen transformation is important for two main reasons: environmental and health hazards. Increased N (along with phosphorus) leads to the natural ecosystem response manifested as eutrophication of surface water bodies, endangering the survival of aquatic biota. Increased nitrogen concentrations in
Data Loading...
