Chemometric approach for prediction of uranium pathways in the soil
No Thumbnail
Authors
Stojanović, Mirjana
Pezo, Lato

Mihajlović, Marija

Petrović, Jelena

Petrović, Marija

Šoštarić, Tatjana

Milojković, Jelena

Article (Published version)

Metadata
Show full item recordAbstract
Understanding the effect of soil parameters (pH, Eh and organic and inorganic ligands availability) on uranium mobility under different geochemical conditions is fundamental for reliable prediction of its behaviour and fate in the environment. In this study, the impact of total and available phosphorus content, humus and acidity of Serbian agricultural soils on the content of total and available uranium were evaluated by Response Surface Methodology (RSM), second order polynomial regression models (SOPs) and artificial neural networks (ANNs). The performance of ANNs was compared with the performance of SOPs and experimental results. SOPs showed high coefficients of determination (0.785-0.956), while ANN model performed high prediction accuracy: 0.8893-0.904. According to the results, total and available uranium content in the soil were mostly affected by pH, statistically significant at p LT 0.05 level. For the same responses the total phosphorus was found to be also very influential..., statistically significant at p LT 0.05 and p LT 0.10 levels. The impact of available phosphourus and humus was much more influential on total and available uranium content, compared to total phosphorus content. Proposed chemometric approach will be very helpful in preserving the natural resources and practical application for risk assessment modeling of uranium environmental pathways.
Keywords:
Uranium / Soil / Response Surface Methodology (RSM) / networks (ANNs) / Chemometric / Artificial neuralSource:
Radiochimica Acta, 2014, 102, 8, 701-710Publisher:
- Walter De Gruyter Gmbh, Berlin
Funding / projects:
- Development of technologies and products based on mineral raw materials and waste biomass for protection of natural resources for safe food production (RS-31003)
- Osmotic dehydration of food - energy and ecological aspects of sustainable production (RS-31055)
DOI: 10.1515/ract-2013-2178
ISSN: 0033-8230
WoS: 000339660900004
Scopus: 2-s2.0-84905220893
Collections
Institution/Community
Institut za tehnologiju nuklearnih i drugih mineralnih sirovinaTY - JOUR AU - Stojanović, Mirjana AU - Pezo, Lato AU - Mihajlović, Marija AU - Petrović, Jelena AU - Petrović, Marija AU - Šoštarić, Tatjana AU - Milojković, Jelena PY - 2014 UR - https://ritnms.itnms.ac.rs/handle/123456789/324 AB - Understanding the effect of soil parameters (pH, Eh and organic and inorganic ligands availability) on uranium mobility under different geochemical conditions is fundamental for reliable prediction of its behaviour and fate in the environment. In this study, the impact of total and available phosphorus content, humus and acidity of Serbian agricultural soils on the content of total and available uranium were evaluated by Response Surface Methodology (RSM), second order polynomial regression models (SOPs) and artificial neural networks (ANNs). The performance of ANNs was compared with the performance of SOPs and experimental results. SOPs showed high coefficients of determination (0.785-0.956), while ANN model performed high prediction accuracy: 0.8893-0.904. According to the results, total and available uranium content in the soil were mostly affected by pH, statistically significant at p LT 0.05 level. For the same responses the total phosphorus was found to be also very influential, statistically significant at p LT 0.05 and p LT 0.10 levels. The impact of available phosphourus and humus was much more influential on total and available uranium content, compared to total phosphorus content. Proposed chemometric approach will be very helpful in preserving the natural resources and practical application for risk assessment modeling of uranium environmental pathways. PB - Walter De Gruyter Gmbh, Berlin T2 - Radiochimica Acta T1 - Chemometric approach for prediction of uranium pathways in the soil EP - 710 IS - 8 SP - 701 VL - 102 DO - 10.1515/ract-2013-2178 UR - conv_694 ER -
@article{ author = "Stojanović, Mirjana and Pezo, Lato and Mihajlović, Marija and Petrović, Jelena and Petrović, Marija and Šoštarić, Tatjana and Milojković, Jelena", year = "2014", abstract = "Understanding the effect of soil parameters (pH, Eh and organic and inorganic ligands availability) on uranium mobility under different geochemical conditions is fundamental for reliable prediction of its behaviour and fate in the environment. In this study, the impact of total and available phosphorus content, humus and acidity of Serbian agricultural soils on the content of total and available uranium were evaluated by Response Surface Methodology (RSM), second order polynomial regression models (SOPs) and artificial neural networks (ANNs). The performance of ANNs was compared with the performance of SOPs and experimental results. SOPs showed high coefficients of determination (0.785-0.956), while ANN model performed high prediction accuracy: 0.8893-0.904. According to the results, total and available uranium content in the soil were mostly affected by pH, statistically significant at p LT 0.05 level. For the same responses the total phosphorus was found to be also very influential, statistically significant at p LT 0.05 and p LT 0.10 levels. The impact of available phosphourus and humus was much more influential on total and available uranium content, compared to total phosphorus content. Proposed chemometric approach will be very helpful in preserving the natural resources and practical application for risk assessment modeling of uranium environmental pathways.", publisher = "Walter De Gruyter Gmbh, Berlin", journal = "Radiochimica Acta", title = "Chemometric approach for prediction of uranium pathways in the soil", pages = "710-701", number = "8", volume = "102", doi = "10.1515/ract-2013-2178", url = "conv_694" }
Stojanović, M., Pezo, L., Mihajlović, M., Petrović, J., Petrović, M., Šoštarić, T.,& Milojković, J.. (2014). Chemometric approach for prediction of uranium pathways in the soil. in Radiochimica Acta Walter De Gruyter Gmbh, Berlin., 102(8), 701-710. https://doi.org/10.1515/ract-2013-2178 conv_694
Stojanović M, Pezo L, Mihajlović M, Petrović J, Petrović M, Šoštarić T, Milojković J. Chemometric approach for prediction of uranium pathways in the soil. in Radiochimica Acta. 2014;102(8):701-710. doi:10.1515/ract-2013-2178 conv_694 .
Stojanović, Mirjana, Pezo, Lato, Mihajlović, Marija, Petrović, Jelena, Petrović, Marija, Šoštarić, Tatjana, Milojković, Jelena, "Chemometric approach for prediction of uranium pathways in the soil" in Radiochimica Acta, 102, no. 8 (2014):701-710, https://doi.org/10.1515/ract-2013-2178 ., conv_694 .