Tidal inlets along the central coast of Vietnam are located in a microtidal, wave-dominated coastal environment. In addition, the Vietnam coast is highly influenced by the seasonal monsoon regime, which is characterized by large northeast waves from October to March and calm southeast waves from April to September every year. Consequently, the tidal inlet entrance morphologies often suffer from a dynamic seasonal evolution due to distinct differences in the direction of wave-induced longshore sediment transport (LST) between the two monsoon seasons. The migration or closure of tidal inlets causes a lot of problems for socio-economic development in the region since these are the main reasons leading to an increase in the risk of coastal flooding and the obstruction of navigation. This paper presents a comprehensive study of the morphological evolutions of natural tidal inlets on the central coast of Vietnam using long-term remote sensing data sets and by the Delft3D numerical model. Surprisingly, the estimated LST rates from the former method are in an order of magnitude agreement with the results from the latter one for all of the areas in this study. Based on the conservation equation for sand and comprehensive data collection, a new simple empirical formula for predicting the sand spit elongation rate as a function of the sand spit width is developed. Although the breaching of sand spit might happen during an extreme flood event at some tidal inlets, the growth rate of the spit before and after the breaching is almost unchanged. These findings are very useful information for supporting the local coastal authorities to find better management solutions in terms of sustainable development.

The potential wave power has been assessed based on long-term wave data along a marginal sea area offshore Phu Yen province in Central Vietnam. Based on the publicly available WaveWatch-III reanalysis wave data (NOAA), the deep-water wave climate during the period from 1989 to 2019 has been analysed and used as the boundary condition for the MIKE21 spectral wave model. The hydrodynamic module of MIKE21 is also run in coupled mode. The model has been calibrated and verified against the measured data at three wave gauges. Simulation has been performed for every month, each with 1-2 typical wave conditions. The results show that the highest wave power (~29 kW/m) occurs in December. The distribution of wave power along the 30-m depth contour has also been presented for the annual average, NE monsoon (winter) average, and S monsoon (summer) average. The distribution map shows that wave power is slightly higher in the south of this area, and the NE monsoon season comes along with much higher wave power (7.4 times compared to that of the S monsoon season). These findings may aid in planning effective exploitation of wave energy for the region.

Land cover and plant species identification using satellite images and deep learning approaches have recently been a widely addressed area of research. However, mangroves, a specific species that have significantly declined in quantity and quality worldwide despite their numerous benefits, have not been the subject of attention. The novelty of this research is to deal with this species based on an advanced deep learning solution (a proposed ensemble U-Net model) and a high-resolution Planet satellite imagery (5 m x 5 m) in a case study of Ngoc Hien district, Ca Mau province, Vietnam. Twelve single U-Net backbone models were trained, and three quantitative metrics (Intersection over Union, F₁-score, and Overall Accuracy) were used to evaluate. The findings indicate that three out of twelve models (MobileNet, SEResNeXt-101 and Efficientnet-B7) experienced the most efficient assessment results for identifying all classes, in which the MobileNet model was the best. These models were applied for the ensemble model’s development. The ensemble model’s quantitative assessment metrics increased considerably by about 3–10% compared to the single-component models. The IoU, F₁-score, and OA values of this model were 80.08%, 95.82%, and 95.90%, respectively. Three classes of mangrove species (Avicennia alba, Rhizophora apiculate, and mixed mangroves) in the ensemble model had more uniform assessment results. In conclusion, to achieve optimal classification outcomes, a land-cover map comprising mangrove species is possibly established using the proposed ensemble model, while a distribution map of mangrove species enables to be developed using the MobileNet model.