Novel Ensemble of MCDM-Artificial Intelligence Techniques for Groundwater-Potential Mapping in Arid and Semi-Arid Regions (Iran)

Abstract

The aim of this research is to introduce a novel ensemble approach using Vise Kriterijumska Optimizacija I Kompromisno Resenje (VIKOR), frequency ratio (FR), and random forest (RF) models for groundwater-potential mapping (GWPM) in Bastam watershed, Iran. This region suffers from freshwater shortages and the identification of new groundwater sites is a critical need. Remote sensing and geographic information system (GIS) were used to reduce time and financial costs of rapid assessment of groundwater resources. Seventeen physiographical, hydrological, and geological groundwater conditioning factors (GWCFs) were derived from a spatial geo-database. Groundwater data were gathered in field surveys and well-yield data were acquired from the Iranian Department of Water Resources Management for 89 locations with high yield potential values ≥ 11 m3 h−1. These data were mapped in a GIS. From these locations, 62 (70%) were randomly selected to be used for model training, and the remaining 27 (30%) were used for validation of the model. The relative weights of the GWCFs were determined with an RF model. For GWPM, 220 randomly selected points in the study area and their final weights were determined with the VIKOR model. A groundwater potential map was created by interpolating the values at these points using Kriging in GIS. Finally, the area under receiver operating characteristic (AUROC) curve was plotted for the groundwater potential map. The success rate curve (SRC) was computed for the training dataset, and the prediction rate curve (PRC) was calculated for the validation dataset. Results of RF analysis show that land use and land cover, lithology, and elevation are the most significant determinants of groundwater occurrence. The validation results show that the ensemble model had excellent prediction performance (PRC = 0.934) and goodness-of-fit (SRC = 0.925) and reasonably high classification accuracy. The results of this study could aid management of groundwater resources and assist planners and decision makers in groundwater-investment planning to achieve sustainability.

Machine Learning-Based Gully Erosion Susceptibility Mapping: A Case Study of Eastern India

“…The multi-collinearity test is an important way to judge the linear dependency among the selected independent factors in the statistical modeling [44]. In the case of the machine learning models, this technique needs to be used for better results [45][46][47][48][49][50][51][52]. Researchers have applied multi-collinearity analysis for gully erosion susceptibility mapping [53], groundwater potentiality mapping [54], landslide susceptibility mapping [48] etc.…”

CMIP5 climate projections and RUSLE-based soil erosion assessment in the central part of Iran

“…Many of these studies integrated remote sensing, Geographic Information System (GIS) and the RUSLE approach for the estimation of SE 50 – 55 . Other recent techniques such as Artificial Neural Networks or Machine Learning techniques are also becoming popular for erosion simulation and modelling in Iran 29 , 56 , 57 . However, despite the numerous studies on SE estimation, there is limited information on SE estimation based on future climate change (CC) scenarios in Iran and other arid and semi-arid countries.…”

Spatial prediction of shallow landslide: application of novel rotational forest-based reduced error pruning tree

“…More recently, machine-learning (ML) ensemble models have been used to map natural hazards. The ML models such as random forest (RAF), boosted regression tree (BRT), artificial neural network (ANN), multivariate adaptive regression spline (MARS), J48 decision tree (JDT), least squares support vector machines (LSSVM), linear discriminant analysis (ADA), decision tree (DT), adaptive neuro-fuzz inference system (ANFIS), k-nearest neighbour (KNN), logistic model tree (LMT), alternate decision tree (ADTree), Bayesian logistic regression (BLR), support vector machine (SVM), convolutional neural network (CNN), and recurrent neural network (RNN) were used in LSM studies by Youssef et al (2016), Zhou et al (2018), Chen et al (2018), Ghorbanzadeh et al (2019) Ngo et al (2021, and Arabameri et al (2019aArabameri et al ( , 2019bArabameri et al ( , 2019cArabameri et al ( , 2019dArabameri et al ( , 2020aArabameri et al ( , 2020bArabameri et al ( , 2020cArabameri et al ( , 2020d. ML models are superior to statistical methods as they are more accurate, have no overfitting problems, and can analyze both continuous and categorical data simultaneously.…”