Quantifying the groundwater total contamination risk using an inclusive multi-level modelling strategy

Multi-hazard assessment for flood and Landslide risk in Kalimantan and Sumatra: Implications for Nusantara, Indonesia's new capital

Situated within the Ring of Fire and characterized by a tropical climate and high seismic activity, Indonesia is uniquely vulnerable to natural disasters such as floods and landslides. These events pose significant threats to both the population and infrastructure. This study predicts areas exposed to flood and landslide risk by considering various environmental factors related to climate, topography, and land use. The predictive performance of three machine learning models-naïve Bayes, k-nearest neighbors, and random forest (RF)-was evaluated by comparing the AUC, RMSE, and R2 values of each model. Ultimately, the RF model, which demonstrated the highest accuracy, was used to prioritize disaster impact factors and generate hazard maps. The results identified the interaction of rainfall, land use, and slope aspect as the most critical determinants of hazard occurrence. The predicted hazard maps revealed that approximately 26.7 % of the study area was vulnerable to either floods or landslides, with 16.8 % of the area experiencing both. The new capital of Nusantara showed a relatively higher multi-hazard risk than did the overall study area and protected zones, with 22.1 % of the hazard area vulnerable to both flooding and landslides. In single hazard zones, areas classified as at risk for floods had a higher mean probability of experiencing both hazards (43 %), as compared to areas classified as at risk for landslides (22 %). As a result, urban planners and relevant stakeholders can now utilize the hazard maps developed in this study to prioritize infrastructure reinforcement and disaster risk areas, integrating land use planning with risk assessment to mitigate the impact of disasters. By employing these strategies, Indonesia and other countries facing similar challenges can now enhance their disaster preparedness and response capabilities in new capital regions and other areas, ultimately planning for more sustainable urban development.

Application and innovation of artificial intelligence models in wastewater treatment

At present, as the problem of water shortage and pollution is growing serious, it is particularly important to understand the recycling and treatment of wastewater. Artificial intelligence (AI) technology is characterized by reliable mapping of nonlinear behaviors between input and output of experimental data, and thus single/integrated AI model algorithms for predicting different pollutants or water quality parameters have become a popular method for simulating the process of wastewater treatment. Many AI models have successfully predicted the removal effects of pollutants in different wastewater treatment processes. Therefore, this paper reviews the applications of artificial intelligence technologies such as artificial neural networks (ANN), adaptive network-based fuzzy inference system (ANFIS) and support vector machine (SVM). Meanwhile, this review mainly introduces the effectiveness and limitations of artificial intelligence technology in predicting different pollutants (dyes, heavy metal ions, antibiotics, etc.) and different water quality parameters such as biochemical oxygen demand (BOD), chemical oxygen demand (COD), total nitrogen (TN) and total phosphorus (TP) in wastewater treatment process, involving single AI model and integrated AI model. Finally, the problems that need further research together with challenges ahead in the application of artificial intelligence models in the field of environment are discussed and presented.

Watershed groundwater level multistep ahead forecasts by fusing convolutional-based autoencoder and LSTM models

The development of deep learning-based groundwater level forecast models can tackle the challenge of high dimensional groundwater dynamics, predict groundwater variation trends accurately, and manage groundwater resources effectively, thereby contributing to sustainable water resources management. This study proposed a novel ConvAE-LSTM model, which fused a Convolutional-based Autoencoder model (ConvAE) and a Long Short-Term Memory Neural Network model (LSTM), to provide accurate spatiotemporal groundwater level forecasts over the next three months. The HBV-light and LSTM models are chosen as benchmarks. An ensemble of point data and the corresponding derived images concerning the past (observations) and the future (forecasts from a conceptual model) of groundwater levels at 33 groundwater wells in Jhuoshuei River basin of Taiwan between 2000 and 2019 constituted the case study. The findings showcase the effectiveness of the ConvAE-LSTM model in extracting crucial features from both point and imagery datasets. This model successfully establishes spatiotemporal dependencies between regional images and groundwater level data over diverse time frames, leading to accurate multi-step-ahead forecasts of groundwater levels. Notably, the ConvAE-LSTM model exhibits a substantial improvement, with the R-squared values showing an increase of more than 18%, 22%, and 49% for the R1, R2, and R3 regions, respectively, compared to the HBV-light model. Additionally, it outperforms the LSTM model in this regard. This study represents a noteworthy milestone in environmental modeling, offering key insights for designing sustainable groundwater management strategies to ensure the long-term availability of this vital resource.

Assessing vulnerability of coastal aquifer to seawater intrusion using Convolutional Neural Network

This study examined coastal aquifer vulnerability to seawater intrusion (SWI) in the Shiramin area in northwest Iran. Here, six types of hydrogeological data layers existing in the traditional GALDIT framework (TGF) were used to build one vulnerability map. Moreover, a modified traditional GALDIT framework (mod-TGF) was prepared by eliminating the data layer of aquifer type from the GALDIT model and adding the data layers of aquifer media and well density. To the best of our knowledge, there is a research gap to improve the TGF using deep learning algorithms. Therefore, this research adopted the Convolutional Neural Network (CNN) as a new deep learning algorithm to improve the mod-TGF framework for assessing the coastal aquifer vulnerability. Based on the findings, the CNN model could increase the performance of the mod-TGF by >30 %. This research can be a reference for further aquifer vulnerability studies.

The assessment of groundwater quality through the water quality and nitrate pollution indexes in northern Türkiye

Groundwater is contaminated by anthropogenic factors such as industry, domestic waste, and excessive fertilizers. Groundwater samples, which were obtained from 50 different wells in July 2020, were used in this study. Thirteen hydrochemical properties, including electrical conductivity (EC), pH, total dissolved solids (TDS), total hardness (TH), nitrate NO3-, anions, and cations were analyzed. Also, types of groundwater were investigated via the Piper diagram. The groundwater was also evaluated for irrigation suitability using the sodium percentage (Na%), sodium adsorption ratio (SAR), Kelly's index (KI), residual sodium carbonate (RSC), potential salinity, magnesium hazard (MR), and permeability index (PI). The samples were assessed for drinking the suitability using the water quality index (WQI) and the nitrate pollution index (NPI). Geographic information systems (GIS) were used to create spatial distribution maps of irrigation water quality indices, WQI, and NPI values. The results of major cations varied sodium 28.69-211.80 mg/L, calcium 78.74-258.89 magnesium 27.78-161.30 mg/L, and potasium 0.10-3.57 mg/L. The results from the study area showed that 62.70 of EC, 32.40% of PI, 20.09% of RSC, 51.55% of PS, and 49.36% of MR were inappropriate for irrigation purposes. The NPI data ranged from - 0.75 to 9.65, and 21.06% of the study areas were heavily polluted. The WQI showed that almost 62.90% of the experimental area was categorized as poor, very poor, and inappropriate for drinking water purposes, whereas 37.10% of the areas were categorized as good and excellent.