Surface sediment classification using a deep learning model and unmanned aerial vehicle data of tidal flats.

This study proposes a deep learning model, U-Net, to improve surface sediment classification using high-resolution unmanned aerial vehicle (UAV) images. We constructed training datasets with UAV images and corresponding labeling data acquired from three field surveys on the Hwangdo tidal flat. The labeling data indicated the distribution of surface sediment types. We compared the performance of the U-Net model trained in various implementation environments, such as surface sediment criteria, input datasets, and classification models. The U-Net trained with five class criteria-derived from previous classification criteria-yielded valid results (overall accuracy:65.6 %). The most accurate results were acquired from trained U-Net with all input datasets; in particular, the tidal channel density caused a significant increase in accuracy. The accuracy of the U-Net was approximately 20 % higher than that of other classification models. These results demonstrate that surface sediment classification using UAV images and the U-Net model is effective.

Depositional environment classification based on environmental variables of South Korea's Nakdong River Estuary barrier-lagoon system.

Anthropogenic activities have altered the geomorphological and ecological conditions of the Nakdong River Estuary (NRE) dramatically over the last century. The objectives of this study were to classify NRE sub-environments and to identify their unique ecological functions. The first step in classification was the establishment of 14 a priori sub-environmental groups based on landscape factors. Surface sediments obtained for these groups were analyzed for factors related to grain size and organic matter. Based on the results, the NRE was divided into two primary estuarine environment divisions: (1) a mixed marine and terrestrial environment influenced primarily by land; and (2) a principally marine environment influenced primarily by the ocean. Using multivariate analysis, we subdivided these primary estuarine divisions into six sub-environments, including Sub-1 and -2 in the former and Sub-3, -4, -5, and -6 in the latter.

Flux and distribution of methane (CH4) in the Gunsan Basin of the southeastern Yellow Sea, off the Western Korea.

The flux and distribution of methane (CH4) was investigated in the seawater column at 14 stations in the Gunsan Basin, the southeastern part of Yellow Sea from 2013 to 2015. Here CH4 is concentrated 2.4-4.7 (3.4 ± 0.7) nM in the surface and 2.5-7.4 (5.2 ± 1.7) nM in the bottom layer. The CH4 saturation ratios ranged from 65.5% to 295.5% (162.6 ± 68.7), comprising the mean sea-to-air CH4 flux of 3.8 to 25.3 (15.6 ± 5.5) µM m-2d-1. Methane concentration was largely different in the upper and the lower seawater layers that is separated by the thermocline of which depth is variable (20-60 m) depending on the time of sampling. The concentration of seawater dissolved CH4 is high between the bottom surface of the thermocline layer and the sea floor. Generally it tends to decrease from the south-westernmost part of the basin toward the west coast of Korea. This distribution pattern of CH4 seems to result from the CH4 supply by decomposition of organic matters produced in the upper seawater layer that is superimposed by the larger supply from the underlying sediment layer especially beneath the thermocline. The latter is manifested by ubiquitous CH4 seeps from the seafloor sediments.

Removal of non-aqueous phase liquids (NAPLs) from TPH-saturated sandy aquifer sediments using in situ air sparging combined with soil vapor extraction.

Contamination in coastal aquifer plains is of great concern in many countries given that non-aqueous phase liquids (NAPLs) have polluted numerous sites through accidental oil spills or improper disposal. We have developed a method to remove pollutants such as NAPLs from sandy sediment samples collected from the Mandol area of Gomso Bay in western South Korea. The sediments were collected from around the diffuser in a two-dimensional (2D) acrylic reaction apparatus, and these contained a total petroleum hydrocarbon (TPH) concentration of 89.3 ppm (mg/kg media). The maximum perchloroethylene (PCE) concentration was 398.51 ppm in the unsaturated zone and 0.77 ppm in the saturated zone. Volatile organic compounds (VOCs) were detected between 20 and 44 hour. However, non-volatile contaminants remained in the sediments after treatment. In situ air sparging (IAS) combined with soil vapor extraction (SVE), transformation from sorbed and nonaqueous phases to the vapor phase, is incomplete when treatment is performed using a pervasive air flow for sediments such as the sand of Mandol. During air transformation, the groundwater flow conditions increased the rate of contaminant removal. Although pilot-scale testing in the field site was fluctuated due to the heterogeneous of sediments condition, this 2D study found that the proposed method can alter the measurable geophysical properties of NAPLs. These findings demonstrate that IAS combined with SVE in the saturated zone is an effective technology for aquifer remediation high applicability of sandy coastal sediments contaminated by NAPLs.

The radius of influence of a combined method of in situ air sparging and soil vapor extraction in the intertidal sediments of Gomso Bay on the west coast of South Korea.

BACKGROUND: In situ air sparging (IAS) was undertaken at sites in the tidal flats of Mandol and Hajeon, on the west coast of South Korea, to estimate variations in the radius of influence (ROI). RESULTS: The Mandol core sample consisted of sand and muddy sand 1.6-3.4 [Formula: see text] (average 2.3 [Formula: see text]) and contained water (average 15.10 %). The Hajeon core sample consisted of muddy sand, sandy silt, and muddy sandy gravel 1.31-4.44 [Formula: see text] (average 3.11 [Formula: see text]) and contained water (average 19.77 %). These sites differ in their sedimentary and geochemical characteristics. At the Mandol site, no H2S or combustible gas was detected during a 48-h sampling period, except for some volatile organic compounds (0.1-2.0 ppm) at the monitoring well during the initial 30 min. At the soil vapor extraction wells, CO2 and O2 varied by 850 ppm (690-1540 ppm) and 0.5 % (20.4-20.9 %), respectively. At the Hajeon site, CO2 and O2 varied from 580 to 1250 ppm and 20.6 to 20.9 %, respectively, during the 48-h sampling period. CONCLUSIONS: At the Mandol site, an oxygen concentration of 20.6 % was assumed as the effective concentration, and the ROI was estimated to be 128.0 cm. However, at the Hajeon site the ROI was estimated to be 85.7 cm. The smaller effective ROI at the Hajeon site was likely caused by the thin aquifer and thin screens of the sparing well. This estimated ROI show that the remediation effectiveness varies greatly as a heterogeneities and anisotropies in the porous sediments. Besides, injection pressure, flow rate, pulsing or continuous mode, and the range of intrinsic permeability for most important characteristic of sediment (soil) type impacted the ROI. Therefore, the IAS method is more effective at a pervasive air flow sediments such as Mandol, which consists of sand and muddy sand than at a channelized site such as Hajeon.

Satellite-based observations of unexpected coastal changes due to the Saemangeum Dyke construction, Korea.

Spatial and temporal changes around an area of conventional coastal engineering can be easily observed from field surveys because of the clear cause-and-effect observable in the before and after stages of the project. However, it is more difficult to determine environmental changes in the vicinity of tidal flats and coastal areas that are a considerable distance from the project. To identify any unexpected environmental impacts of the construction of Saemangeum Dyke in the area, we examined morphological changes identified by satellite-based observations through a field survey on Gomso Bay tidal flats (15km from Saemangeum Dyke), and changes in the suspended sediment distribution identified by satellite-based observations through a hydrodynamic analysis in the Saemangeum and Gomso coastal area. We argue that hydrodynamic changes due to conventional coastal engineering can affect the sedimentation pattern in the vicinity of tidal flats. We suggest that the environmental impact assessment conducted before a conventional coastal engineering project should include a larger area than is currently considered.

Decalcification of benthic foraminifera due to "Hebei Spirit" oil spill, Korea.

In order to determine the effects on foraminifera due to spilled crude oil in the "Herbei Spirit" incident, a study of benthic foraminiferal assemblages was carried out on sediment samples collected from the Sogeunri tidal flat, Taean Peninsula, Korea. Breakages of the chambers in the Ammonia beccarii and Elphidium subincertum species of the Sogeunri tidal flat with a low pH (6.98 on average) were marked. These chamber breakages occurred in 71.6% of A. beccarii and are thought to be caused by decalcification due to the fall in pH resulting from the "Hebei Spirit" oil spill. The factors that affect breakage of the chamber in benthic foraminifera under low pH condition may be not only deto decalcification but also to exposure duration of substrata in the tidal flat spilled crude oil.

Macrobenthos habitat potential mapping using GIS-based artificial neural network models.

This paper proposes and tests a method of producing macrobenthos habitat potential maps in Hwangdo tidal flat, Korea based on an artificial neural network. Samples of macrobenthos were collected during field work, and eight control factors were compiled as a spatial database from remotely sensed data and GIS analysis. The macrobenthos habitat potential maps were produced using an artificial neural network model. Macrobenthos habitat potential maps were made for Macrophthalmus dilatatus, Cerithideopsilla cingulata, and Armandia lanceolata. The maps were validated by compared with the surveyed habitat locations. A strong correlation between the potential maps and species locations was revealed. The validation result showed average accuracies of 74.9%, 78.32%, and 73.27% for M. dilatatus, C. cingulata, and A. lanceolata, respectively. A GIS-based artificial neural network model combined with remote sensing techniques is an effective tool for mapping the areas of macrobenthos habitat potential in tidal flats.