Application of Soft Computing Models for Simulating Nitrate Contamination in Groundwater: Comprehensive Review, Assessme

  • PDF / 1,890,789 Bytes
  • 23 Pages / 595.276 x 790.866 pts Page_size
  • 102 Downloads / 204 Views

DOWNLOAD

REPORT


ORIGINAL PAPER

Application of Soft Computing Models for Simulating Nitrate Contamination in Groundwater: Comprehensive Review, Assessment and Future Opportunities Masoud Haghbin1 · Ahmad Sharafati2,3,4   · Barnali Dixon5 · Vinod Kumar6 Received: 5 April 2020 / Accepted: 12 October 2020 © CIMNE, Barcelona, Spain 2020

Abstract Groundwater is one of the major resources to supply the agriculture and urban water demand. Vulnerability of groundwater resources due to chemical substances is a crucial concern for groundwater quality management. The different nitrogen compounds, especially nitrate, plays an important role in groundwater quality. In last two decades, the efficient approaches called soft computing (SC) models were used for assessing the groundwater pollution. This study aims to assess the applications of various SC models for simulating the groundwater pollution due to nitrate contamination. In this way, the past trends and current applications of those models and essential factors required for assessing the ground water quality are demonstrated. Ultimately, several research gaps and possible future research direction are proposed.

1 Introduction Human activities affect the quality of groundwater resources due to harmful chemical substances produced by agricultural and industrial practices. Nitrate is the most ubiquitous chemical pollutant existed in many aquifers around the world [51]. Nitrate has great potential to penetrate into the groundwater and thus, a high concentration of nitrate (more than 10 mg/ lit) is harmful due to adverse impact on human health [61]. Assessing the groundwater pollution (GWP) due to nitrate contamination is a challenging issue because of complexity of nitrate transport and related uncertainties [76]. * Ahmad Sharafati [email protected]; [email protected] 1



Young Researchers and Elites club, Science and Research Branch, Islamic Azad University, Tehran, Iran

2



Institute of Research and Development, Duy Tan University, Da Nang 550000, Vietnam

3

Faculty of Civil Engineering, Duy Tan University, Da Nang 550000, Vietnam

4

Department of Civil Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran

5

Geospatial Analytics Lab, School of Geosciences, University of South Florida, Tampa, USA

6

Department of Botany, Government Degree College Ramban, Jammu, India



Due to rapid advancement in field of computer science in past two decades, a new field called soft computing (SC) was developed as an alternative method for modeling the GWP [41]. The SC approaches provide an acceptable performance in different field of engineering [71]. Several SC techniques such as fuzzy logic (FL)-based models, artificial neural network (ANN), support vector machines (SVM), self-organized mapping (SOM), decision trees (DT) and random forest (RF) have provided a high performance to map a significant relation between the target and input variables [41]. Earliest application of SC methods for assessing the pollution of groundwater due to nitrate contaminati