Flood-prone area mapping using machine learning techniques: a case study of Quang Binh province,Vietnam

Natural Hazards - Vietnam’s central coastal region is the most vulnerable and always at flood risk, severely affecting people’s livelihoods and socio-economic development. In...     

Identification of groundwater mean transit times of precipitation and riverbank infiltration by two‐component lumped parameter models

Groundwater transit time is an essential hydrologic metric for groundwater resources management. However, especially in tropical environments, studies on the transit time distribution (TTD) of groundwater infiltration and its corresponding mean transit time (mTT) have been extremely limited due to data sparsity. In this study, we primarily use stable isotopes to examine the TTDs and their mTTs of both vertical and horizontal infiltration at a riverbank infiltration area in the Vietnamese Mekong Delta (VMD), representative of the tropical climate in Asian monsoon regions. Precipitation, river water, groundwater, and local ponding surface water were sampled for 3 to 9 years and analysed for stable isotopes (δ¹⁸O and δ²H), providing a unique data set of stable isotope records for a tropical region. We quantified the contribution that the two sources contributed to the local shallow groundwater by a novel concept of two‐component lumped parameter models (LPMs) that are solved using δ¹⁸O records. The study illustrates that two‐component LPMs, in conjunction with hydrological and isotopic measurements, are able to identify subsurface flow conditions and water mixing at riverbank infiltration systems. However, the predictive skill and the reliability of the models decrease for locations farther from the river, where recharge by precipitation dominates, and a low‐permeable aquitard layer above the highly permeable aquifer is present. This specific setting impairs the identifiability of model parameters. For river infiltration, short mTTs (<40 weeks) were determined for sites closer to the river (<200 m), whereas for the precipitation infiltration, the mTTs were longer (>80 weeks) and independent of the distance to the river. The results not only enhance the understanding of the groundwater recharge dynamics in the VMD but also suggest that the highly complex mechanisms of surface–groundwater interaction can be conceptualized by exploiting two‐component LPMs in general. The model concept could thus be a powerful tool for better understanding both the hydrological functioning of mixing processes and the movement of different water components in riverbank infiltration systems.     

Environmental health impacts of mobility and transport in Hai Phong, Vietnam

Traffic is an essential part of modern society and mobility is part of its socio-economic setting. However, signs of counter productivity arise as the current mobility patterns substantially affect our health, including the consequences from traffic accidents, air pollution—which causes even more victims than traffic accidents—and traffic-noise. The use of private motorised vehicles also contributes to sedentarism, climate change and psychological effects. This paper reviews these mobility related health effects and applies them to the situation in Hai Phong, a Vietnamese port-city in fast development and facing growing mobility patterns. In his Master Plan the city developed a view on its development together with the transportation infrastructure up to 2020. Together with the fast changing mobility patterns, such as a modal change from bicycles to motorcycles and cars, this lead to an increase in motorized vehicles and non-negligible environmental health risks. Applying the methodology of a Health Impact Assessment as used in previous studies the current health burden is estimated, focussing on particulate matter (PM) and noise. For PM₁₀ 1287 deaths per year were calculated for the current situation, where the estimated number of deaths by 2020 doubled up to 2741. Hospital admissions due to PM₁₀ raised from 44,954 now to 51,467 in 2020 and for PM_2.5 the restricted-activity days were calculated, accounting for 852,352 per year. For noise only calculations for the current state (2007) were performed. The total estimated DALYs due to noise was 4758.     

Predicting water flow in fully and partially saturated porous media: a new fractal-based permeability model

Hydrogeology Journal - Predicting the permeability of porous media in saturated and partially saturated conditions is of crucial importance in many geo-engineering areas, from water resources to...     

Improved flood susceptibility mapping using a best first decision tree integrated with ensemble learning techniques

Improving the accuracy of flood prediction and mapping is crucial for reducing damage resulting from flood events. In this study, we proposed and validated three ensemble models based on the Best First Decision Tree (BFT) and the Bagging (Bagging-BFT), Decorate (Bagging-BFT), and Random Subspace (RSS-BFT) ensemble learning techniques for an improved prediction of flood susceptibility in a spatially-explicit manner. A total number of 126 historical flood events from the Nghe An Province (Vietnam) were connected to a set of 10 flood influencing factors (slope, elevation, aspect, curvature, river density, distance from rivers, flow direction, geology, soil, and land use) for generating the training and validation datasets. The models were validated via several performance metrics that demonstrated the capability of all three ensemble models in elucidating the underlying pattern of flood occurrences within the research area and predicting the probability of future flood events. Based on the Area Under the receiver operating characteristic Curve (AUC), the ensemble Decorate-BFT model that achieved an AUC value of 0.989 was identified as the superior model over the RSS-BFT (AUC = 0.982) and Bagging-BFT (AUC = 0.967) models. A comparison between the performance of the models and the models previously reported in the literature confirmed that our ensemble models provided a reliable estimate of flood susceptibilities and their resulting susceptibility maps are trustful for flood early warning systems as well as development of mitigation plans.     

The occurrence of cyanobacteria and microcystins in the Hoan Kiem Lake and the Nui Coc reservoir (North Vietnam)

The occurrence of increasing blooms of toxic cyanobacteria in freshwaters has received much attention due to the ability of many cyanobacteria to produce potent cyanotoxins. In this paper, the occurrence of dominant cyanobacteria and the concentration of microcystins (MCs) analysis were investigated monthly from July 2008 to April 2009 in the Hoan Kiem Lake and from February to April 2009 in the Nui Coc reservoir. Concentrations of intracellular MCs from water, bloom samples, and isolated strains were quantified by using high performance liquid chromatography (HPLC). During the study period, the microscopic examination of the phytoplankton samples showed the dominance of the genus Microcystis in the water environment of the Hoan Kiem Lake and the Nui Coc reservoir. The toxin analysis by HPLC demonstrated the presence of two MC variants: MC-LR and MC-RR in water samples. Total concentrations of the toxins in filtered samples from surface water ranged from non-detected to 0.91 μg L^?1 at Nui Coc reservoir and they ranged from 2.1 to 46.0 μg L^?1 at Hoan Kiem Lake. The results of the HPLC analysis confirmed the production of MCs in bloom samples (ranged from 115.9 to 184.6 μg L^?1 in the Hoan Kiem Lake and from 726.5 to 1116 μg L^?1 in the Nui Coc reservoir) and isolated strains of Anabaena sp. and Microcystis with the concentration of MC ranging from 152 to 396.2 μg g^?1 dry mass, respectively.     

Multi-hazard assessment using machine learning and remote sensing in the North Central region of Vietnam

Natural hazards constitute a diverse category and are unevenly distributed in time and space. This hinders predictive efforts, leading to significant impacts on human life and economies. Multi-hazard prediction is vital for any natural hazard risk management plan. The main objective of this study was the development of a multi-hazard susceptibility mapping framework, by combining two natural hazards—flooding and landslides—in the North Central region of Vietnam. This was accomplished using support vector machines, random forest, and AdaBoost. The input data consisted of 4591 flood points, 1315 landslide points, and 13 conditioning factors, split into training (70%), and testing (30%) datasets. The accuracy of the models' predictions was evaluated using the statistical indices root mean square error, area under curve (AUC), mean absolute error, and coefficient of determination. All proposed models were good at predicting multi-hazard susceptibility, with AUC values over 0.95. Among them, the AUC value for the support vector machine model was 0.98 and 0.99 for landslide and flood, respectively. For the random forest model, these values were 0.98 and 0.98, and for AdaBoost, they were 0.99 and 0.99. The multi-hazard maps were built by combining the landslide and flood susceptibility maps. The results showed that approximately 60% of the study area was affected by landslides, 30% by flood, and 8% by both hazards. These results illustrate how North Central is one of the regions of Vietnam that is most severely affected by natural hazards, particularly flooding, and landslides. The proposed models adapt to evaluate multi-hazard susceptibility at different scales, although expert intervention is also required, to optimize the algorithms. Multi-hazard maps can provide a valuable point of reference for decision makers in sustainable land-use planning and infrastructure development in regions faced with multiple hazards, and to prevent and reduce more effectively the frequency of floods and landslides and their damage to human life and property.     

Seasonal variability of cohesive sediment aggregation in the Bach Dang–Cam Estuary, Haiphong (Vietnam)

In the Bach Dang–Cam Estuary, northern Vietnam, mechanisms governing cohesive sediment aggregation were investigated in situ in 2008–2009. As part of the Red River delta, this estuary exhibits a marked contrast in hydrological conditions between the monsoon and dry seasons. The impact on flocculation processes was assessed by means of surveys of water discharge, suspended particulate matter concentration and floc size distributions (FSDs) conducted during a tidal cycle at three selected sites along the estuary. A method was developed for calculating the relative volume concentration for the modes of various size classes from FSDs provided by the LISST 100X (Sequoia Scientific Inc.). It was found that all FSDs comprised four modes identified as particles/flocculi, fine and coarse microflocs, and macroflocs. Under the influence of the instantaneous turbulent kinetic energy, their proportions varied but without significant modification of their median diameters. In particular, when the turbulence level corresponded to a Kolmogorov microscale of less than ∼235 μm, a major breakup of flocs resulted in the formation of particles/flocculi and fine microflocs. Fluctuations in turbulence level were governed by seasonal variations in freshwater discharge and by the tidal cycle. During the wet season, strong freshwater input induced a high turbulent energy level that tended to generate sediment transfer from the coarser size classes (macroflocs, coarse microflocs) to finer ones (particles/flocculi and fine microflocs), and to promote a transport of sediment seawards. During the dry season, the influence of tides predominated. The turbulent energy level was then only episodically sufficiently high to generate transfer of sediment between floc size classes. At low turbulent energy, modifications in the proportions of floc size classes were due to differential settling. Tidal pumping produced a net upstream transport of sediment. Associated with the settling of sediment trapped in a near-bed layer at low turbulent energy, this causes the silting up of the waterways leading to the harbour of Haiphong.