Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T., 2020. Special Issue on “Advances in Geospatial Research of Coastal Environments”. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. vi-xiii. Coconut Creek (Florida), ISSN 0749-0208.

Geospatial research in the fields of remote sensing (RS), geospatial information system (GIS), global positioning system (GPS), digital photogrammetry (DP) has become essential to understanding the coastal environments. Thus, to create a multidisciplinary forum of discussion on recent advances in geospatial research of coastal environments, original research articles and literature review papers addressing advances in geospatial research of the coastal environments have been considered for the publication in this special issue. Finally, a total of 40 papers was published in this special issue. In this editorial paper, we review the previous special issue related to geospatial research of coastal environments and summarize the papers published in this special issue.

Wang, J.B.; Ren, G.B.; Lin, Z.Y.; Wang, A.D.; Hu, Y.B.; Li, X.M.; Wu, P.Q.; Ma, Y., and Zhang, J., 2020. Estimation and analysis of nitrogen contents in the yellow river estuary wetland using Gaofen-1 remote sensing data. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 1-10. Coconut Creek (Florida), ISSN 0749-0208.

Nitrogen storage is a key indicator of vegetation growth, and the spatial distribution of nitrogen storage provides important information about the ecological status of wetland ecosystems. In this study, the aboveground vegetation nitrogen content (AVNC) in the Yellow River estuary wetland is estimated based on GaoFen-1 wide field of view (GF-1 WFV) image. Based on the spatial distribution of vegetation and the fractional vegetation coverage (FVC), the characteristics of AVNC are analyzed. On the basis of this study, the best model for estimating AVNC based on vegetation indices is the sample rate index (SRI), whose determination coefficient (R²) is 0.72 and root mean square error (RMSE) is 1.5 g/m². Per the results of this model, the total AVNC of the study area is obtained, which is 620.6 tons.

The total ANVC for freshwater Phragmites australis (FPA), Tamarix chinesis shrub (TCS), Spartina alterniflora (SA), Suaeda salsa (SS), and tidal flat Phragmites australis (TFPA) communities are 296.2 t, 115.3 t, 97.1 t, 86.3 t and 25.7 t, respectively. The average AVNC per unit area for SA, FPA, TCS, TFPA, and SS communities are 4.2 g/m², 4.8 g/m², 3.1 g/m², 1.7 g/m² and 1.4 g/m², respectively. The correlation between FVC and AVNC is also analyzed, from which, the spatial distribution of AVNC and FVC is found to be in good agreement. Results of this study thus provide a valuable reference for determining AVNC in coastal wetland ecosystems with implications for ecological health and other ecological characteristics.

Wan, J.X. and Ma, Y., 2020. Multi-scale spectral-spatial remote sensing classification of coral reef habitats using CNN-SVM. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 11-20. Coconut Creek (Florida), ISSN 0749-0208.

In recent years, coral reefs have undergone serious degradation globally, prompting the use of remote sensing as an effective means to monitor these reefs on a large scale. Deep learning, which is a state-of-the-art image processing technique suitable for remote-sensing applications, can be used to learn nonlinear characteristics of images and is therefore applicable to the classification of small-scale coral reefs. This paper proposes a multi-scale method based on a convolutional neural network and support vector machine (CNN-SVM) to classify the coral reef habitats of Zhaoshu Island and Zhong Island in the Xisha Archipelago, China. This method combines spectrum, texture, and bidimensional empirical mode decomposition (BEMD) based scale separation algorithm to fully learn multi-scale information of coral reefs. Remote-sensing images captured by the WorldView-2 and Gaofen-2 (GF-2) satellites are used to evaluate the performance of the proposed CNN-SVM framework. The results indicate that the proposed method performs accurately and efficiently. Compared with SVM, random forest (RF), CNN, and CNN-RF, the overall accuracy is improved by 10.57 %, and the accuracy of classifying reef-clumping areas is improved by 17.44 %.

Li, X.M.; Ma, Y.; Leng, Z.H.; Zhang, J., and Lu, X.X., 2020. High-accuracy remote sensing water depth retrieval for coral islands and reefs based on LSTM neural network. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 21-32. Coconut Creek (Florida), ISSN 0749-0208.

Accurate water depth data are essential to ensure navigational safety of ships operating in regions with coral islands and reefs; however, it is often difficult or impossible to conduct in situ bathymetric surveys in such areas. In this study, a Long-Short Term Memory (LSTM) neural network in deep learning was introduced for multispectral remote sensing detection of water depth, and an LSTM neural network model suitable for remote sensing water depth retrieval for coral islands and reefs was developed. The LSTM model retrieval result was optimal when using an Adam optimizer, batch size of 10 %, 2000 epochs and a 50/100 network structure. Compared with the classical Log-linear, Stumpf and improved Stumpf models, the LSTM model demonstrated better water depth retrieval capability with minimum mean absolute errors (MAEs) and mean relative errors (MREs) irrespective of whether using 300 or 2000 training points. The retrieval accuracy of the LSTM model with only 300 training points was better than the other three models with 2000 training points. Compared with SVR model, MRE of LSTM model reduced 19.03 % and 4.14 % respectively when using 300 and 2000 training points. Analysis of the Dong Island case showed that the water depth retrieval results based on the LSTM model clearly reflected the geomorphic units of the entire reef such as the reef flat, front slope (seaward slope), and patch reef. It also revealed subtle geomorphologic features of the reef flat surface and the valley system. The LSTM model demonstrated satisfactory performance for water depths of 5–25 m, with average MREs of 7 % and 9 % (minimum: 4.02 %, maximum: 14.62 %). These findings verify the application performance of the LSTM remote sensing water depth retrieval model both qualitatively and quantitatively.

Han, H.G. and Lee, M.J., 2020. A method for classifying land and ocean area by removing sentinel-1 speckle noise. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 33-38. Coconut Creek (Florida), ISSN 0749-0208.

In Korea, satellite image-based land cover maps are limited because they are based on time-consuming pixel value-based classification techniques. The main categories of land cover classification are water and land; therefore, synthetic-aperture radar (SAR) images with high water reflectivity may be used to improve land cover map classification accuracy. In this study, C-band SAR images obtained by the Sentinel-1 satellite are used, which include various noises including speckle noise. To remove speckle noise, this paper applied Lee, Gamma, and Frost filters, and found that the Lee filter offered the best performance. By combining a stacking technique and the Lee filter, this paper successfully classified the target water system using image dichotomy and histogram analyses of the region of interest (ROI). The resulting land cover map showed 90 % accuracy compared to the pixel-based map, and comparison with the optical image showed that water coverage was effectively classified. This classification of forest reservoirs, which are difficult to distinguish in optical images, was rated as excellent. Thus, this speckle noise-removal technique will facilitate the improvement of land cover classification accuracy, particularly for flood boundaries and shorelines.

Lee, S.M.; Lee, E.J.; Yoo, H.S., and Lee, M.J., 2020. Analysis of trends in marine water quality using environmental impact assessment monitoring data: a case study of Busan new port. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 39-46. Coconut Creek (Florida), ISSN 0749-0208.

Various development projects, including harbor construction, are carried out in coastal areas. In South Korea, the performance of environmental impact assessment (EIA) is legally required for development projects in marine areas to monitor the impacts of these projects on offshore marine environments. National Marine Environmental Measurement Network (NMEMN) data can be used to examine the overall status of the marine environment prior to a development project, but these data are very limited. In contrast, EIA data are collected from a large number of locations over various periods. The purpose of this study was to analyze trends in the coastal marine environment using NMEMN and EIA data collected in 2010–2011 for Busan New Port, where multiple development projects are underway. Use of the combined dataset enabled more effective analysis of trends in marine water quality, including those in dissolved oxygen, the chemical oxygen demand (COD), total nitrogen (TN), and total phosphorus (TP), than did the use of NMEMN data alone. Values increased in the bottom layer, especially in the summer, after port operation had begun. For example, average summer ranges for COD, TN, and TP in the bottom layer in 2000–2005 were 1.41–2.2 mg/L, 0–0 µg/L, and 0–17.62 µg/L, respectively, and those in 2006–2011 were 1.53–3.2 mg/L, 0–281 µg/L, and 8.09–49.35 µg/L, respectively. In the future, data from various sources should be integrated into EIAs to identify not only the impacts of a single project, but also the overall impacts of multiple changes in an area. This method also could be used to provide basic data for continuous management of the marine environment.

Cho, N.W. and Lee, M.J., 2020. Exploring changes in coastal environment policy using text mining: A case study in South Korea. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 47-53. Coconut Creek (Florida), ISSN 0749-0208.

Land development activities performed in coastal areas are accompanied by environmental impacts, and Environmental Impact Assessment (EIA) is used to mitigate such impacts. The characteristics of coastal environmental policy can be determined based on the content of EIA reports for development projects However, owing to the vast amounts of information regarding various environmental factors, including air, water, and soil, contained in such reports, their potential for analysis is limited. This study analyzed the terms of agreement for development projects—the result of EIA of coastal areas—using bibliometrics, a method that is widely used to analyze trends in academic research. Data on the EIA of coastal development projects conducted in South Korea over the past 25 years were collected, and the surveyed period was sub-divided into five-year periods to build datasets through text mining. Subsequently, trends in coastal environmental policy were analyzed over an extended period. The results of the analysis indicated that keywords related to the water environment showed the highest appearance frequency for the entire period due to the characteristics of coastal development projects. Keywords related to the natural ecological environment continued to show an increasing trend, while those related to the land environment showed a continuously decreasing trend. Based on correspondence analysis, the five sub-divided periods were broadly classified into three groups, and the principal inertia of the analysis was found to be approximately 89.6 %. The characteristics of the coastal environmental policy that were emphasized for each period were determined; land and life environments were found to be emphasized for the period between 1994 and 1999; water environments were emphasized between 2000 and 2009; and natural ecology environments were emphasized between 2010 and 2019. By contrast, atmospheric and socio-economic environments did not show a significant correlation. This study proposes methods of utilizing information provided by EIA reports to analyze environmental policy and offers new insights specifically applicable to coastal environmental policy.

Zhu, J.S.; Hu, P.; Zhao, L.L.; Gao, L.; Qi, J.W.; Zhang, Y., and Wang, R.F., 2020. Determine the Stumpf 2003 model parameters for multispectral remote sensing shallow water bathymetry. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 54-62. Coconut Creek (Florida), ISSN 0749-0208.

The Stumpf 2003 model is a widely used model for shallow water bathymetry estimation for multi-spectral remote sensing. There are three parameters should be determined in this model. One is the parameter n in the term of a logarithmic ratio, and the other two parameters are m₀ and m₁, which are determined by a linear regression. In many researches the n is assigned as a constant directly. There is no more discussion on how to determine these parameters, especially for the value of n. In this paper, it suggest a two-step-method when using the Stumpf 2003 model for bathymetry estimation. The first step determines the value of the parameter n according to the linearity between the term of a logarithmic ratio and water depth. The second step is to obtain the other two parameters m₀ and m₁ by a conventional linear regression. The method is tested and verified using a WorldView-2 (WV2) multi-spectral image and the corresponding in-situ water bathymetry by a sonar. 512 samples are extracted randomly as a train dataset, these data are used to train the Stumpf 2003 model; the remain 129 samples are collected as a validation dataset. The results show that using the two-step-method can improve the accuracy of bathymetry estimation. According to the train dataset, the RMSE is 3.829 using the model parameters (n=54.766) determined by this method, while the RMSE is 4.005 using the model parameters (n=1000) determined in the conventional way. Similar RMSE results also obtained for the validation dataset, they are 1.753 (n =54.766) and 1.816 (n=1000) respectively. It also shows an improvement. So this paper suggest to use the two-step-method to determine the model parameters for bathymetry estimation when using Stumpf 2003 model.

Kim, S,-K.; Kim, J.-C., and Choi, J.-Y., 2020. The effect of laver-farming on the distribution of copepod community in the west coast of Korea. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 63-68. Coconut Creek (Florida), ISSN 0749-0208.

To understand the effect of farms on copepod communities in the coastal marine environment, this study investigated the environmental variables and copepod communities in June 2016 in the coastal region adjacent to the Geum River estuary. Copepod communities were more abundant in laver farms than in areas not covered by laver farms, due to the availability of various food sources and low predation risks. Although laver farms had low abundance (cells/mL) of phytoplankton, they were observed to have high species diversity. This provides an environment in which a variety of food items is available for copepod communities. Moreover, fishes were more abundant in the areas not occupied by farms than in laver farms. the complex habitat structure created by abundant laver, as well as frequent human interference in laver farms, has a negative impact on the distribution of fish communities. However, the abundance of copepod communities in laver farms is important for a continuous supplement to the diet of fishes in the marine environment, where food sources are likely to be depleted. Thus, the positive effects of laver farms on copepod communities can greatly contribute to the population growth of fishes and to the sustainability of copepod communities in the coastal marine environment. However, the artificial disturbances that frequently occur in laver farms are likely to lead to eutrophication and various types of water pollution in the coastal marine environment, so proper control and management of human activity are required.

Choi, J.-Y.; Kim, J.-C., and Kim, S.-K, 2020. Changing distributions of zooplankton communities in a coastal lagoon in response to rainfall seasonality. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 69-74. Coconut Creek (Florida), ISSN 0749-0208.

Coastal lagoons are unique and complex ecosystems which exhibit very high primary and secondary production rate, and their productivity is underpinned by plankton communities. Empirical studies suggest that summer rainfall determines the spatial and temporal pattern of not only physicochemical variables but also zooplankton communities in coastal regions. It was hypothesized that drier or wetter than average summer rainfall years would change the environmental variables (e.g., salinity, dissolve oxygen, and nutrients) in coastal lagoons, thereby disrupting the spring growth trend of zooplankton communities. Long-term (2014–2017) monitoring data (rainfall, environmental variables, and zooplankton) were divided into two groups: Rainy and Dry years, corresponding to years with an annual rainfall that was higher or lower than the total annual average, respectively. The results showed that summer and autumn densities of zooplankton fell sharply in Rainy years but increased steadily in Dry years. The highest density of rotifers was mainly observed in sites adjacent to the inflow of tributary streams, while copepods were abundant near the outlet to the ocean. The differing spatial distribution of rotifers and copepods is attributed to the salinity gradient in the study site, and changed with freshwater inflows. Based on these findings, it is suggested that summer rainfall variations play an important role in driving the spatial and temporal distribution of zooplankton.

Hakim, W.L.; Achmad, A.R.; Eom, J., and Lee, C.-W., 2020. Land subsidence measurement in Jakarta coastal area using time series interferometry with Sentinel-1 SAR data. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 75-81. Coconut Creek (Florida), ISSN 0749-0208.

Jakarta coastal area is a part of North Jakarta which has a boundary to the Java Sea. Coastal flood from the Java Sea was not a new case in North Jakarta, the situations occurred due to sea-level rise. The conditions of coastal flood in Jakarta can be worsening by land subsidence in Jakarta coastal area. The land subsidence in Jakarta coastal area was occurred in many years and detected by geodetic measurement methods such as GPS and leveling surveys. The study about land subsidence in Jakarta coastal area needed to be updated because it could affect the urban development. This study presents the mean vertical deformation map using Stanford Methods for Persistent Scatterer (StaMPS) for time-series analysis on the Sentinel-1 data archives in the period of March 2017 through April 2020 in both ascending and descending track. The comparison of mean vertical deformation map between two tracks shows a good correlation in terms of displacement patterns with slightly different in 4 individual chosen points that show a high deformation at rates between 30 mm/year and 40 mm/year. The area chosen in this study known as a reclaimed area and the subsidence occurred due to young alluvium consolidation that not supported the maximum compression from a lot of buildings. These approach methods in this study that using StaMPS could be used to monitor land subsidence from all-terrain.

Zhao, S.; Xu, Y.; Li, W., and Lang, H., 2020. Optical remote sensing ship image classification based on deep feature combined distance metric learning. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 82-87. Coconut Creek (Florida), ISSN 0749-0208.

Image classification of ships has become an active research topic in the field of remote sensing. Effectively distinguishing specific category of a ship is crucial for many maritime matters. At present, there are two major difficulties in ship classification tasks. Firstly, the construction of traditional artificial features not only requires the assistance of professional knowledge, but also fails to represent the rich semantic information of ship images. Secondly, the classification of ships requires the identification of subcategory, the problem of intra-class diversity and inter-class similarity has become more serious and requires more careful handling. In order to solve the above problems, this paper proposes a method combining the deep feature and the distance metric learning (DML) algorithm to implement the classification of ships in the optical remote sensing images. In this study, an improved fine-tuned AlexNet convolution neural network (CNN) is adopted to extract the deep feature of ship images. The DML algorithm is applied to learn a good distance metric to reveal the similarity and dissimilarity between ships which are represented by deep feature. Comprehensive experiments show that the classification performance of deep feature combined DML algorithm significantly outperforms the comparative methods.

Kim, T.; Kim, Y.-S., and Lee, H.-J., 2020. Characteristics of geological lineaments along the eastern coast of the Korean Peninsula: A statistical approach. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 88-100. Coconut Creek (Florida), ISSN 0749-0208.

Lineament analysis is an essential method for evaluation of site safety in important facilities such as nuclear power plant and nuclear waste disposal site. However, in the case of coastal areas, if the same criteria used for inland is applied, large-scale lineament could be underestimated. Therefore, a different classification criterion is necessary to evaluate the characteristics of lineament pattern along coastal areas. Thus, here, a new statistical method focusing on zonal analysis is proposed. This method is applied for evaluating the characteristics of lineaments developed within 5 km from the eastern coastline of the Korean Peninsula. The most dominant direction of lineaments along the eastern coast lies in the N–S direction, followed by the NE–SW, and the E–W directions. Since the length distribution of the lineament shows the characteristics of the lognormal distribution, statistical analysis was performed using probability density functions (PDFs). The results indicate that the mode value of the lineament length is about 2 km, and lineaments shorter than 6 km account for 90 % of the total. To confirm the validity of the statistical distribution features, re-analyzing the study area by progressively increasing its width to 10 km, 15 km, 20 km, and 30 km was done. The analysis showed that the distribution characteristics were similar to those of the area with a width of 5 km. The statistical analysis using PDFs improved the existing classification criteria based on lineament length and provided three new quantitative classes of lineaments within coastal regions.

Kim, J.C.; Lee, H.J.; Choi, J.Y.; Kim, S.K. and Jung, H.S. 2020. Topographical change in coastal areas arising from soil erosion in the Riparian zone. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 101-106. Coconut Creek (Florida), ISSN 0749-0208.

Sedimentary inputs into coastal estuaries because of activities upstream can lead to rapid bathymetric and topographical changes, which may affect coastal functions such as habitat or fisheries. In South Korea, the Four Major Rivers Restoration Project was initiated in 2008 to address environmental problems arising from dredging and land reclamation, and a number of ecological parks were created. This study analyzed: 1) the land cover of each period using land cover maps before (2007), immediately after the project (2012), and after five years (2017); 2) the average annual amount of soil loss from the riparian zone of the Nakdong River using the revised universal soil loss equation (RUSLE); and 3) the change in the area of barrier islands in the estuary of the Nakdong River from 2008 to 2017. In order to generate the factors required for RUSLE, various spatial data, such as land cover maps, national spatial information, aerial photographs, a soil map, and average annual precipitation data were used. The results showed the percentage of annual soil loss from classes 4, and 5 (severe and very severe levels of soil erosion, respectively) increased immediately after the project finished in 2012 but decreased by 2017. The digitized aerial photographs confirmed that the area of Maenggeummeori-do, one of the barrier islands located in the estuary of the Nakdong River, decreased by more than 20 % as sediment inputs declined. Thus, it was confirmed that changes in soil loss in riparian zones arising from river maintenance projects can affect the area of barrier islands in coastal estuaries. Through this study, it was revealed that large-scale river refurbishment projects indirectly affected soil erosion in the waterfront area of the Nakdong River, which eventually affected the area change of the barrier island at the mouth of the Nakdong River.

Kim, S.-K.; Kim, J.-C., and Choi, J.-Y., 2020. Distribution and utilization plan of reclaimed lands (RLs) as waterfowl habitat on the south coast of Korea. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 107-112. Coconut Creek (Florida), ISSN 0749-0208.

Since 1960, in South Korea, many coastal areas have been reclaimed to secure agricultural land and construct levees. Small-scale reclaimed lands (RLs) were isolated from coastal areas by levees and roads, and this resulted in the loss of the intrinsic characteristics of the ocean environment. In this study, the distribution of RLs located on the south coast of South Korea was investigated, and the formation cause and use types were identified. Furthermore, their potential was assessed as an alternative wetland by investigating waterfowl distribution in each RL. A total of 161 RLs (total area: 776 ha) were located near the southern coastal area of South Korea. Most RLs were created through disconnection from the marine environment due to the construction of an embankment or road, and most of the land was left unattended because it was difficult to use as agricultural or residential areas due to its narrowness and intermittent water level. The majority of RLs were characterized by intermitted water levels, deep water depth, and high salinity, making it difficult for plants other than emergent plants to develop, and consequently, fauna, including birds, were unable to utilize the environment as a habitat. In winter, some RLs were utilized as habitats for ducks (i.e., dabbling ducks and diving ducks), but there were few in the following seasons (spring and autumn). However, an RL was judged to have potential as a wetland if ecological function improvements, such as efficient water supply and water depth control, could be made. In particular, increasing the diversity and coverage rate of aquatic plants can lead to abundant food sources, such as fish and invertebrates. The use of these RLs as wetlands not only enhances the value of abandoned RLs but can also be a significant alternative habitat for diverse animals at a time when inland wetlands are being damaged and reduced.

Zhang, X.; Zhu, Y. X.; Zhang, J.; Meng, J.; Li, X., and Li, X., 2020. An algorithm for sea ice drift retrieval based on trend of ice drift constraints from Sentinel-1 SAR data. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 113-126. Coconut Creek (Florida), ISSN 0749-0208.

Realizing high-resolution (sub-kilometer scale) automatic detection of sea ice drift from massive Synthetic Aperture Radar (SAR) data is an important requirement for mastering polar sea ice dynamics and better understanding global climate change. A fully automatic SAR sea ice drift detection method has been developed in this paper, which is based on the idea of multi-scale observation. First, ice cracks or lead structures was detected from down-sampled sequential SAR images. Because the ice cracks or leads have scattering intensities and shape features which differ significantly from the surrounding ice regions, these structures can help to detect the overall drift of the sea ice region which we called the trend of ice drift (TID). Then, a modified feature tracking (FT) method was used to extract sea ice drift vectors from the original SAR imagery by considering the constraints of TID. The aims of the proposed algorithm are to improve the accuracy of sea ice drift retrieval and the spatial density of ice drift vectors, as well as reducing execution time. Three pairs of dual-polarization Sentinel-1 SAR images of the Arctic were used for experimental validation. The experimental results show that the proposed method takes the least time, and the extracted sea ice drift vector is far more than the classical Speeded-up Robust Features (SURF) method and Normalized cross-correlation (NCC) method in both spatial density and area coverage. In terms of the accuracy of sea ice drift speed and direction detection, for HH polarization, the sea ice drift speed retrieved by the proposed method has a Root Mean Square Error (RMSE) of 0.158 cm/s and a relative error (RE) of 1.838 %; and the RMSE of retrieved sea ice drift direction is 0.112° and the RE is 0.267 %. For HV polarization, the RMSE of sea ice drift velocity inversion is 0.138 cm/s and RE is 1.504 %; the RMSE of sea ice drift direction inversion is 0.123°and RE is 0.753 %. For the dual-polarization SAR data, compared with SURF and NCC methods, the proposed algorithm performs better.

Chun, J.-H.; Kim, Y.; Choi, J.-Y.; Kim, Y.J.; Lee, H.-J., and Kim, Y.H., 2020. Observation of diffusive methane emission from Holocene mud deposits on the continental shelf offshore southeastern Korea. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 127-136. Coconut Creek (Florida), ISSN 0749-0208.

This study investigated gas-charged sediments found in Holocene mud deposits on the inner continental shelf offshore southeastern Korea. Headspace gases obtained from seven piston cores of Holocene mud deposits were analyzed using gas chromotography, and methane dissolved in the water column were analyzed using in situ methane sensors and gas chromotography. The methane concentration data from both sediments and water columns provide the evidences of diffusive methane emission from the inner continental shelf. The methane concentration of headspace gases ranged from 22.2 to 123.9 mM below the upper front of acoustic blanking zone (ABZ). The maximum anomaly of dissolved methane in the water column was 38.3 nM at a site 5 over rough seafloor in the ABZ, which is surrounded by water masses of relatively lower methane concentration. Dissolved methane concentrations were significantly lower (< 10 nM) in the upper water column (< 20 m water depth). Concentrations of dissolved methane within the whole water column in the gas-free zone (GFZ) were < 5 nM. These results imply that diffusive methane emission occurred from a rough seafloor with erosional features on the inner continental shelf at water depth of about 54 m; this diffusive methane emission was not related to high methane concentrations in piston core sediments of the shallow upper front of the ABZ.

Baek, W.-K.; Jung, H.-S.; and Kim, D., 2020. Oil spill detection of Kerch strait in November 2007 from dual-polarized TerraSAR-X image using artificial and convolutional neural network regression models. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 137-144. Coconut Creek (Florida), ISSN 0749-0208.

The oil spill is the main marine disaster. It is known that the data mining based method performs better in detecting oil than the traditional SAR based method to distinguish from lookalikes. Recently, artificial neural networks were employed to detect the oil spill. For that synthetic aperture radar intensity map, intensity texture map, co-polarized coherence map, co-polarized phase difference texture map, and digital elevation model was employed as input. The detection performance based on data mining largely depends on the used data, the characteristics of the study area, and the used model. Improving detection performance through various existing data mining models is a very important task since minor improvements in detection can have a significant impact on disaster damage mitigation. Artificial and convolutional neural network regression models were applied to detect the oil spill of Kerch strait in November 2007. The two models showed F1-scores of 0.832 and 0.823 respectively. The results of this study would contribute to monitor oil spill dynamic and mitigate damages caused by the oil spill.

Yang, Q.; Wang, G.; Zhang, X.; Grecos, C., and Ren, P., 2020. Coastal image captioning. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 145-150. Coconut Creek (Florida), ISSN 0749-0208.

Coastal images convey immense semantic information of the corresponding coastal areas. This article presents a preliminary approach to coastal image captioning that describes salient semantic information of coastal images with accurate and meaningful sentences. Specifically, this article exploits a state-of-the-art method referred to as self-critical sequence training for coastal image captioning. Firstly, a convolutional neural network (CNN) produces fixed-length features from the coastal images. Secondly, the fixed-length features are fed into a long short-term memory (LSTM) to generate captions, i.e., the accurate and meaningful sentences. Finally, the LSTM is optimized in the context of reinforcement learning. Experimental evaluation on the Moye island remote sensing dataset validates the effectiveness.

Tai, X.; Wang, G.; Grecos, C., and Ren, P., 2020. Coastal image classification under noisy labels. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 151-156. Coconut Creek (Florida), ISSN 0749-0208.

Remote sensing coastal images commonly suffer from noisy labels, which tend to reduce the benefits of training coastal image classifiers and thus deteriorate the classification performance. This article explores the feasibility of classifying coastal images under noisy labels. The state-of-the-art co-teaching paradigm in computer vision is exploited for processing remote sensing coastal images that are partially incorrectly labeled. The data with noisy labels are hard to learn and induce large loss values for a classifier. In the light of this observation, the co-teaching paradigm trains two identically structured but randomly initialized convolutional neural network (CNN) classifiers with stochastic optimization. Training of the two CNN classifiers is performed on mini-batches in an interactive and iterative manner. Specifically, the data with small loss values of one CNN classifier are used for updating the other CNN classifier in the next training iteration. The two CNN classifiers turn robust to noisy labels through the iterative process of teaching each other. Experimental evaluations on a coastal image dataset, i.e., the Moye Island remote sensing dataset, validate the effectiveness of the proposed coastal image classification strategy.

Park, N.-W. and Jang, D.-H., 2020. Geostatistical classification of intertidal surface sediments using log-ratio transformation and high-resolution remote sensing imagery. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 157-165. Coconut Creek (Florida), ISSN 0749-0208.

This paper presents a multivariate geostatistical approach to classify intertidal surface sediments by combining compositional data analysis and high-resolution remote sensing imagery. An isometric log-ratio (ilr) transformation is first applied to the sediment composition data prior to employing geostatistical analysis to consider the compositional properties of the sediment compositions. To complement the information deficiency of sparse field measurements, high-resolution remote sensing imagery is considered as exhaustive soft information and integrated with the ilr transformed balances via simple kriging with varying local means (SKLM). An inverse ilr transformation is then applied to the SKLM results to obtain sediment compositions over the study area. Finally, Shepard's classification scheme is applied to the sediment compositions to classify the intertidal surface sediments. A case study was conducted on the Baramarae tidal flats in Korea with high-resolution KOMPSAT-2 imagery to demonstrate the effectiveness of the proposed geostatistical approach. The classification results produced by the integration of high-resolution remote sensing imagery via ilr transformation and SKLM outperformed those based on cokriging of sediment compositions, with an improvement of approximately 11.7 %p in overall accuracy. This improvement was attributed to the superior prediction capability of SKLM for sediment compositions. Further, detailed variations in the sediment distributions in the study area, which could not be observed when using only a limited number of sediment samples, could be revealed by integrating the high-resolution remote sensing imagery. Therefore, the geostatistical integration approach that properly accounts for both the property of sediment compositions and the exhaustive soft information from remote sensing imagery could be effectively applied for the classification of intertidal surface sediments.

Niu, L. and Lang, H., 2020. Ship matching using convolutional neural network in multi-source synthetic aperture radar images. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 166-175. Coconut Creek (Florida), ISSN 0749-0208.

There are two major challenges in the ship matching and tracking task by synthetic aperture radar (SAR) images. First, images from different satellites have different incidence angles, polarization modes and resolutions, leading to the nonlinear geometric deformation of ships and the difference in image quality. The above differences result in a significant drop in the similarity of matching ships. Second, the principal direction of ships is arbitrary, which requires that the matching method has rotation-invariance. For the first challenge, the siamese-type convolution neural networks are adopted to reduce the distance between matching ships and increase the distance between non-matching ships. For the second challenge, a minimum bounding square (MBS) segmentation method is used to process all the ship images to the horizontal direction. After image preprocessing using MBS, there are only two situations for the principal direction of ships and rotation-invariance is transformed into the invariance of prow-to-stern interchange. To solve the problem of prow-to-stern interchange, a constrained hinge loss is proposed to ensure that the minimum similarity of matching images is still greater than the maximum similarity of non-matching images. There are a variety of siamese-type convolutional neural networks with good performance currently, the focus of the research is to use the two proposed strategies (MBS and constrained hinge loss) to improve the performance of existing methods instead of designing network architectures. In order to verify the feasibility of the proposed strategies, a multi-source ship target database covers different SAR sensors is constructed for ship matching. The experimental results prove that the proposed methods can match the ship target precisely and outperform the methods without MBS preprocessing and the constrained hinge loss function.

Cao, M.; Qing, S.; Du, Y.; Yuan, R.; Shun, B.; Hao, Y., and Zhao, W., 2020. Remote sensing classification of aquatic vegetation in Ulansuhai Lake based on discrete particle swarm optimization algorithm. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 176-186. Coconut Creek (Florida), ISSN 0749-0208.

As the largest natural wetland in the same latitude on the earth, water ecological environment of Ulansuhai Lake, China is seriously threatened, and there is various aquatic vegetation spread over the lake. Remote sensing is considered to be an effective way to map the distribution of aquatic vegetation. Different from the method used in most of the previous studies, discrete particle swarm optimization (DPSO) algorithm was used to identify and classify emergent vegetation (EV), yellow algae (YA), submerged aquatic vegetation (SAV) and water in Ulansuhai Lake based on Landsat-8 Operational Land Imager (OLI) in this research. The classification results were validated by 284 investigation sites data and visual interpretation of Gao Fen 2 (GF-2) image. The results indicated that determination coefficient (R²) for EV, YA, SAV and water were greater than 0.91, root mean square error (RMSE) were less than 0.025 km². Besides, the overall accuracy (95.4 %) and Kappa coefficient (0.93) of DPSO algorithm are superior to spectral index, unsupervised classification methods and supervised classification methods. In addition, DPSO algorithm to other regions (the Yellow Sea) and sensors (Sentinel-2) have been successfully applied, which further proves the applicability of DPSO algorithm. The research provides a new tool to assist people in locating and quantifying aquatic vegetation, so that purposeful actions can be taken to control the eutrophication of lake water and improve the water ecological environment.

Choi, D.-L.; Shin, D.-H.; Jin, J.-Y.; Lee. Y.-K., and Kum, B.-C., 2020. High-resolution seismic stratigraphy offshore Haeundae beach in Busan, South Korea. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 187-193. Coconut Creek (Florida), ISSN 0749-0208.

The stratigraphic architecture of the Haeundae shallow sea was evaluated using numerous high-resolution single-channel seismic profiles and drilled cores. Two major seismic units over the basement, the lower Unit II and upper Unit I, as well as two subunits (Units Ia and Ib) in the upper Unit I were identified in the study area. Unit II shows numerous chaotic diffraction hyperbolic seismic reflection patterns overlying high amplitude, and irregular reflection of the bedrock surface, corresponding well with the gravel-dominated deposits. Unit II is interpreted as nonmarine alluvial deposits in the Quaternary glaciations. Unit I is present over Unit II with two distinct subunits at different locations: Unit Ia, distributed inshore within ∼10 m depth, and Unit Ib, offshore beyond ∼13 m depth. Unit Ia, which is accumulated more in the western inshore section, exhibits an acoustically semi-transparent and sigmoidal clinoform wedge and is well correlated with sand-dominated deposits from core data. Unit Ib, which pinches out landward but gradually thickens seaward, is generally characterized by transparent seismic facies, likely made up of very fine sandy to muddy sediments from surface sediments. The depositional environments of Units Ia and Ib are categorized as shoreface settings and the landward innermost part of the Nakdong distal delta system during late Holocene sea-level highstand, respectively. Due to severe weather conditions associated with winter storms and summer typhoons as well as human interventions, the sediment morphology and its distribution are thought to have constantly been reactivated and modified.

Li, Y.; Wang, J., and Wang, S., 2020. Absolute salinity measurement based on microfiber coaxial Mach-Zehnder interferometer. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 194-201. Coconut Creek (Florida), ISSN 0749-0208.

Seawater salinity is a key parameter in the study of ocean dynamics, accurate measurement of seawater salinity is crucial. However, current seawater salinity is a measure of practical salinity, which limits the accuracy of salinity measurements. The proposed absolute salinity compensates for the flaws in the practical salinity that does not accurately reflect the true salinity of seawater. Microfiber sensors enable highly sensitive measurements of absolute salinity in seawater. In this paper, a new method based on microfiber coaxial Mach-Zehnder interferometer (coaxial MZI) is proposed for seawater absolute salinity measurement. The absolute salinity sensitivity of microfiber coaxial MZI is calculated by finite element program, and the theoretical results show that the sensitivity increases with increasing wavelength and increases with the decrease of waist diameter. Thirteen microfiber coaxial MZIs with different diameter were fabricated and used for the measurement of absolute salinity. Experimental results show that the highest sensitivity is 2.38nm/ ‰, and the variations rules of sensor sensitivity are consistent with theoretical calculations. In addition, in order to promote the practical application of microfiber devices in the marine field, coaxial MZI is encapsulated. The salinity response time, time stability and vibration influence of the sensor were tested by experiments. Encapsulated coaxial MZI is expected to realize the in-site measurement of the absolute salinity of seawater.

Shin, J.; Kim, S.M., and Ryu, J.-H., 2020. Machine learning approaches for quantifying Margalefinium polykrikoides bloom from airborne hyperspectral imagery. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 202-207. Coconut Creek (Florida), ISSN 0749-0208.

The ichthyotoxic mixotrophic dinoflagellate Margalefidinium polykrikoides is well known for killing fish in aquaculture cages via gill clogging at high cell abundance. In order to establish a countermeasure against M. polykrikoides blooms, it is essential to monitor and quantify the spatial abundance of red tide cells with high accuracy. In this study, machine-learning (ML) approaches were applied to investigate the spectral features of M. polykrikoides cell abundance on the south coast of Korea. Four ML models were developed to retrieve the M. polykrikoides cell abundance in optically complex waters from airborne hyperspectral imagery (HSI). The models included the feed-forward neural network (FFN), support vector machine (SVM), ensemble bagged tree (EBT), and Gaussian process regression (GPR). Paired data of in situ spectra and M. polykrikoides cell abundance were used to train and validate the ML models. The performances of ML models were evaluated using the linear regression value (R²) and root mean squared error (RMSE), considering the predicted cell abundances calculated from HSI and ground truth. The GPR model utilized the validation data and showed the best performance (R² = 0.57 and RMSE = 659.18 cells mL^-1). GPR was also able to estimate the cell abundance with over 3,000 cell mL^-1, indicating viable results. Therefore, using GPR is suggested to obtain more accurate red tide cell abundance maps. Local-based ML models for red tide cell abundance retrieval from HSI data can support the monitoring and surveillance of red tide blooms on the south coast of Korea.

Park, W.; Baek, W.-K.; Won, J.-S., and Jung, H.-S., 2020. Comparison of input image size for ship detection from KOMPSAT-5 SAR image using deep neural networks (DNNs). In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 208-217. Coconut Creek (Florida), ISSN 0749-0208.

This study shows the results of ship detection in KOMPSAT-5 X-band SAR images using Deep Learning (DL) based the fast Deep Neural Network (DNN) method with the iterative kernel-based false alram detection algorithm. In addition, this study verifies the detection accuracy according to the size of the input data for are typical errors in SAR images such as backscattering noise, side-lobe, and ghost effect. The test results of the Fast DNN method confirmed that the error caused by the back scattering noise and side-lobe decreased as the input data size increased, however, the error for the ghost effect did not improve significantly. Also, the False Alarm Rate (FAR) was greatly improved using the iterative kernel-based false alram detection algorithm. However, caution is needed when detecting small ships because very small boats disappear. The results of this study showed that the size of the input data was very effective at 51 × 51 or higher when the fast DNN method and the iterative kernel-based false alram detection algorithm were applied to the KOMPSAT-5 images to detect ships, allowing the FAR and Intersaction over Union (IoU) showing relatively high accuracy at 0.270 and 0.611.

Nur, A.-S.; Achmad, A.-R., and Lee, C.-W., 2020. Land subsidence measurement in reclaimed coastal land: Noksan using C-Band Sentinel-1 radar interferometry. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 218-223. Coconut Creek (Florida), ISSN 0749-0208.

It has been 27 years since the beginning of the Noksan national industrial complex construction at downstream of the Nakdong River in Busan. The complex was built by landfilling a coastal area over one of Korea's deepest soft ground areas. Land subsidence occurred since the start of construction due to primary consolidation and secondary consolidation after the creation finished. Since secondary consolidation directly affects the structure over Noksan, it is crucial to monitor the subsidence rate. This study found that land subsidence happened in four areas of Noksan National Industrial Complex from 2017 to 2020 by applying Stamford Methods for Persistent Scatterer (StaMPS) time-series method using C-band Sentinel-1 SAR datasets from descending and ascending track. The results show land subsidence with a maximum velocity of 20.98 mm/year from descending track datasets and 20.92 mm/year from ascending track datasets. The highest rate area of land subsidence occurred in the south part of Noksan. Mean vertical deformation maps between descending and ascending track have been compared and show a good relationship with 0.93 of coefficient correlation. These approach methods in this study that using StaMPS could be used to monitor slow rate of land subsidence in reclaimed coastal land.

Kim, K.; Kim, B.-J.; Kim, E., and Ryu, J.-H., 2020. Classification of green tide at coastal area using lightweight UAV and only RGB images. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 224-231. Coconut Creek (Florida), ISSN 0749-0208.

Remote sensing has attracted much attention as a realistic solution to monitor green tide outbreaks quickly and efficiently, and the scope of its utilization is also gradually increasing. The use of remote sensing from satellites and airborne platforms has many advantages when monitoring large areas, but its utilization is very limited considering the cost and spatio-temporal resolution. In this work, the availability of high-resolution unmanned aerial vehicle (UAV) image detection for the detection of green tides on the Jeju coast of Korea is presented, and the classification accuracy was evaluated through the application of various classification algorithms. The UAV survey area was 1.0 km², and the spatial resolution of the UAV images taken at an altitude of 250 m was 4.86 cm. In this study, it was divided the survey area into four classes: Ulva, sand, seawater, and submerged Ulva. classifying them was attempted using only the RGB bands of a lightweight UAV. The high-resolution UAV images were classified as the Mahalanobis distance (MHD), maximum likelihood (MLH), minimum distance (MID), and artificial neural network (ANN) algorithms, with the highest accuracy being for the MLH and ANN methods. The green tide was calculated by counting only pixels classified as Ulva and submerged Ulva in the UAV image to determine the area of the green tide in the research area. The green tide areas estimated by the MHD, MLH, MID, and ANN classification algorithms were 0.29, 0.38, 0.30, and 0.37 km², respectively. Given that many unclassified pixels have been found using the MLH method and that this phenomenon has not been found in the ANN method, the ANN algorithm can be considered to be the most effective for coastal green tide classification. This study showed that high-resolution RGB images of lightweight UAVs could produce highly satisfactory classification results. The methods presented in this study can be considered as a very effective approach in terms of their ability to quickly acquire high-resolution images at low cost and detect vegetation with high accuracy. In the future, more diverse targets and areas will be available for identification if an accuracy assessment of classification results is made based on the spatial resolution of UAV images.

Lee, Y.-K.; Choi, J.-K., and Lee, H.-J., 2020. A study on seasonal dynamics of suspended particulate matter in Korea coastal waters using GOCI. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 232-245. Coconut Creek (Florida), ISSN 0749-0208.

Analysis of suspended particulate matter (SPM) is a key to understanding the turbulent sediment flow in the Heuksan mud belt (HMB), located along the southwestern coast of the Korean Peninsula. The purpose of this study was to investigate intra-annual variability in remotely sensed SPM data derived from the Geostationary Ocean Color Imager (GOCI) around the HMB over a period of 1 year (2013). Monthly averaged SPM images obtained by the GOCI showed pronounced seasonal changes in turbid water. Dominant environmental factors for intra-annual and diurnal variability of SPM, including river discharge, winds, and tides, were assessed. Monsoon wind dominated seasonal variation in SPM dispersion; the extent of turbid water increased during the winter and decreased during the summer months. Diurnal variation observed from 1-day composite images showed clearly that tidal current was the dominant factor affecting the short-term dispersal pattern of SPM. The gate-controlled river discharge showed the reverse seasonal pattern compared with SPM concentration. However, a turbid plume from the Geum River observed in GOCI data could support the identification of this river as the major source of the HMB. In conclusion, this study demonstrated that the spatial and temporal dynamics of the HMB, including intra-annual and diurnal variability, can be successfully detected using the GOCI.

Kim, K.-L. and Ryu, J.-H., 2020. Mapping oyster reef distribution using Kompsat-2/3 and linear spectral unmixing algorithm – A case study at Hwangdo tidal flat. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 246-253. Coconut Creek (Florida), ISSN 0749-0208.

Oyster areas are widely used as species of biological indicator, and they are also the most important shellfish in terms of ecosystem economic valuation. Oysters are production on the west coast before 2007 accounted for about 5 % of the nation's total production, but the 2007 Hebei Spirit oil spill caused production to plunge to less than 1 %. Therefore, spatial distribution maps of oyster reefs are required to help local authorities define management strategies. In this study, Kompsat-2/3 was used to map oyster reef distribution and analyze the distribution of oyster reefs based on linear spctral unmixing in Hwangdo tidal flat. A spectral library, collected in situ for various conditions with a field spectroradiometer, was used to conduct linear spectral unmixing classifications on Kompsat-2/3 data. The classification result shows very high accuracy, with an overall accuracy of 95 % or more, and there were misclassifications in some areas. The most causes of misclassification are the similarity of spectral characteristics and the limitations of the spatial and spectral resolutions of satellite images. For this reason, it was difficult to distinguish between oyster reefs and area distributed with many macrobenthos, and small-young oyster reefs were difficult to classify due to very weak reflectivity. In addition, the sand bar, it was is difficult to distinguish between oyster reefs related to dead reefs and sandbars because of the high reflectivity of these areas in the imagery. As a result of analyzing the change in the oyster reef area, it increased in 2019 compared to that in 2015. Especially, the oyster reef area increased in 30-50 % sand content and decreased in 20-30 % and 60-70 %. The changes of sand sediment seem to affect the distribution of oyster reefs. This study may be useful for mapping the distribution of oyster reefs and understanding the spatial variation of their habitat.

Koo, S.M.; Seo, J.M.; Song. Y.J., and Baek, S.J., 2020. Storage class memory based hybrid memory system for practical remote sensing. In: Jung, H.-S.; Lee, S., and Ryu, J.-H. (eds.), In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 254-260. Coconut Creek (Florida), ISSN 0749-0208.

In the remote sensing domain, large-sized, files such as high-resolution satellite images or sonar video clips from unmanned underwater vehicles are very common. For processing big files, a large main memory is necessary. The main memory capacity highly influences the performance of computer systems. The demand for DRAM capacity has never been satisfied, and more importantly, large DRAM systems suffer from significant power consumption. To ease the problem, a promising solution is to build a hybrid main memory (HMM) system composed of a small number of fast DRAMs and many inexpensive devices. By mimicking large and fast main memory capacity, HMM allows computer systems to run applications that require more DRAM than is installed on the system. In this paper, a novel HMM management scheme for a storage class memory (SCM)-based HMM was introduced. As all the data stored on SCM are already non-volatile, the overall performance of the computer system is enhanced further by not flushing them periodically. The proposed idea was implemented on Linux and its performance was measured using an SCM emulation system. It was shown that HMM efficiently improves performance by up to 77.9 %, compared with a conventional operating system. Additionally, it was demonstrated that the proposed idea's fault recovery mechanisms could restore dirty data that are not yet synchronized with storage, within 91 % of the test time.

Kim, I.; Park, S.-I., and Lee, H.-J., 2020. Structural analysis of a fault-related anticline in the southwestern Gyeonggi Massif, Korea using an unmanned aerial vehicle and field surveys: The role of rejoining splays in a duplex-like structure. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 261-270. Coconut Creek (Florida), ISSN 0749-0208.

In fold and thrust belts, various contractional structures accommodate horizontal shortening and vertical thickening in response to orogenic events. In this study, it was conducted a structural analysis of the Gomseom anticline, exposed along the coast in a part of a Phanerozoic orogenic belt in the southwestern Gyeonggi Massif, on the southern Korean Peninsula. To combat difficulties related to accessing the outcrop under tidal effects, high-resolution unmanned aerial vehicle (UAV) images alongside field data was used. The UAV photography technique allowed easy macro and mesoscale structural mapping and down-plunge projection, thereby enabling the efficient interpretation of the geometric and kinematic features of the Gomseom anticline. In the core of the Gomseom anticline, fault-bounded sheets are vertically stacked and folded, giving rise to a unique structural geometry analogous to duplexes. This duplex-like structure is bounded by a submerged thrust at the base and tightly folded thrust imbrication at the top and defined by intervening shortcut rejoining splays branching from the latter. Taking the kinematics of the sheet-bounding faults and intra-sheet minor faults into account, it was suggested an evolutionary model of the Gomseom anticline in which repeated fault-related folding and sequential propagation of rejoining splays give rise to the duplex-like structure. In the rejoining splay model, displacement on the basal thrust is transferred to a passive-roof thrust without hard links between them. Our result can be evaluated as a possible alternative duplex model that contrasts pre-existing models such as “Boyer-type”, “connecting splay”, and “fold” duplexes. However, further testing of the model is required.

Wu, M.; Zhao, Y.F.; Sun, L.E.; Huang, J.; Wang, X.H., and Ma, Y., 2020. Remote sensing of spatial-temporal variation of chlorophyll-a in the Jiaozhou Bay using 32 years Landsat data. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T.W. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 271-279. Coconut Creek (Florida), ISSN 0749-0208.

The chlorophyll-a concentration (Chla, µg/L) is a vital indicator of water quality and eutrophication, yet optical complexity and significant variability of coastal waters make the accurate estimation of Chla challenging. Monitoring spatial-temporal distribution and variation of Chla and comprehending the correlation between Chla and environmental factors are necessary for long-term water quality assessment. This study calibrated and validated the Chla estimation model with satisfactory performance (R², RMSE, and MRE values are 0.77, 0.64 µg/L, and 32.5 %) and further characterized the spatial-temporal variation of Chla in Jiaozhou Bay (JZB) based on 381 cloud-free Landsat images of 32 years (1986-2017). The annual mean Chla in JZB reached the highest value in 1997 and decreased gradually in the following two decades. The seasonal variation of Chla is obvious: the highest value of Chla appeared in summer, followed by spring, autumn and winter. Accordingly, the monthly averaged Chla peaked in July, while the minimum occurred in January. The spatial distribution of Chla on different time scales shared a similar pattern. High Chla appeared in the northwestern part of JZB and decreased gradually to the southwest, resulting in the lowest Chla near the water channel connected to the open sea. The spatial heterogeneity mainly arose from river discharge, while the temporal heterogeneity may be caused by seasonal variations in precipitation, temperature, and river discharge. This study indicated that the empirical models for the Landsat data could effectively monitor the long-term Chla variation in JZB.

Sun, W.; Zhang, J.; Meng, J.; Li, Y., and Cao, K., 2020. An arctic gridded sea surface temperature product constructed from spaceborne radiometer data. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 280-286. Coconut Creek (Florida), ISSN 0749-0208.

A sea surface temperature (SST) record with a grid size of 0.1° for the Arctic region is presented by fusing satellite-derived SST products from three passive-microwave (WindSat, AMSR-E and ASMR2) and four infrared (AVHRR, MODIS onboard Terra and Aqua, and VIIRS) radiometers (RMs). In this study, optimum interpolation (OI) is used to construct the dataset. The fused SST data of Arctic covering the regime north of 60°N span from 2010 to 2018, available with the format of NetcDF. The accuracy of the gridded Arctic SST product was assessed with the drifting buoy measurements. Comparisons with the measurements from Argo buoys show that the Root Mean Squared Errors (RMSE) of the fusion SST data obtained in this study is relatively stable with values ranging from 0.54 to 0.75 °C. This gridded SST data set is characterized with a high spatiotemporal resolution and an assured accuracy.

Zhao, W.; Zhao. G.; Yang, J.; Xu, M., and Kou, J., 2020. Assessment of the ecological security of the coastal zone of the Guangdong/Hong Kong/Macao greater bay area. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 287-295. Coconut Creek (Florida), ISSN 0749-0208.

With the impact of global climate change and the acceleration of urbanization, the coastal zone of the Guangdong/Hong Kong/Macao Greater Bay area (GBA) faces the pressures of biodiversity reduction, loss of coastal wetlands, eutrophication of coastal waters and frequent marine pollution events, all of which affect the ecology of this environment and threaten high-quality development of the GBA coastal zone. This study looked at the coastal zones adjacent to nine cities in the GBA, and analyzed the temporal and spatial changes in key ecological and environmental factors using remote sensing, field observations and data collection. A pressure-state-response model was used to build coastal zone ecological security evaluation indexes. The resulting assessments based on the ecological security index of each administrative region in the GBA were used to evaluate their ecological security status, diagnose their existing problems and analyze the causes in each of them. The results showed that the state grade of the GBA coastal zone follows a V-shape, with the eastern and western parts in better ecological condition than the central part. With regard to the current management of the coastal ecological environment, this paper suggests six future policies for improvements to coastal protection in an effort to promote high-quality development in GBA coastal zones, including strengthening top-level planning, optimizing port design and strengthening supervision.

Wang, C.; Yang, J.; Li, J., and Chu, J., 2020. Deriving natural coastlines using multiple satellite remote sensing images. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 296-302. Coconut Creek (Florida), ISSN 0749-0208.

The exact boundary of a coastline is difficult to define accurately, and it is even more difficult to determine from remote sensing images. One definition of coastlines is the average spring tide lines. However, establishing the position of this time-sensitive location accurately from temporally sparse remote sensing observations is challenging, especially for natural coastlines such as sandy and muddy coasts. This paper proposes a method that uses multiple remote sensing images to derive coastlines under the assumption that shorelines are invariable over short periods. First, instantaneous water edges between water and non-water regions are extracted from multiple remote sensing images. Second, the extracted water edges are clustered into two groups through application of the K-means algorithm. Next, two average water edges are calculated separately from the two groups. Then, in combination with known tidal levels (amplitudes) at the image capture times, a method for deriving coastline is presented under the assumption that the coastal slope is gentle and the water depths of the locations in the same water edges are equal. Five images acquired by the Operational Land Image sensor of Landsat 8 are used to extract a coastline with known tidal amplitudes to validate the method. Experimental results show that our proposed method could derive highly accurate coastlines from multiple remote sensing images for the regions of artificial coasts and muddy coasts (except estuary areas).

Yue, Y.L.; Qing, S.; Diao, R.X., and Hao, Y.L., 2020. Remote sensing of suspended particulate matter in optically complex estuarine and inland waters based on optical classification. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 303-317. Coconut Creek (Florida), ISSN 0749-0208.

Accurate suspended particulate matter (SPM) concentration retrieval across complex estuarine to inland waters from ocean color remote sensing reflectance (R_rs(λ)) faces challenges. In this paper, an optical classification-based SPM retrieval algorithm in optically complex estuarine and inland waters was proposed and tested in the Yellow River Estuary and Daihai Lake, China. Firstly, the in situ measured R_rs(λ) (n = 204) were classified into two optical water types with the method defined by Matsushita et al. (2015). Secondly, we designed several mathematical models and selected the optimal algorithm according to the goodness of fit. Optimal algorithms were developed for each water type to achieve accurate SPM retrieval. Through the construction of the optimal retrieval algorithm in each water type, the uncertainty of SPM retrievals has been reduced from 95 % to about 39 % compared with the algorithm without optical classification. The retrieval algorithm based on optical water classification was further applied to the Sentinel-2 MSI L2A data over the study area and produced reliable SPM maps. Independent validation with the in situ-satellite match-ups further demonstrates the algorithm's validity (uncertainty of about 47 %). In contrast, applications of other SPM retrieval algorithms resulted in less reliable SPM results with either unsatisfactory retrieval accuracy in class1 (the lowest value of r can reach 0.02). The optical classification, together with the optimal retrieval algorithm for each optical type, is proved to be a feasible way for SPM retrieval in high accuracy over optically complex waters.

Yu, H.F.; Wang, C.Y.; Sui, Y.; Li, J.H., and Chu, J.L., 2020. Automatic extraction of green tide using dual polarization China GF-3 SAR images. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T.W. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 318-325. Coconut Creek (Florida), ISSN 0749-0208.

Traditional optical images are not suitable for the all-weather observation of green tide due to the fact that they are greatly affected by clouds. Synthetic aperture radar (SAR) on the other hand, is able to realize all-day and all-weather observation due to a lower sensitivity to clouds, rain and fog. SAR images are therefore an effective supplementary data for the monitoring of green tide. Due to the differences in brightness and noise across the different areas in an image, it is difficult to use the same threshold in order to extract all the green tide information from the image. Based on the iterative threshold method and the histogram bimodal method, this paper presents a new method for the automatic detection of green tide that uses adaptive thresholds for Gaofen-3 (GF-3) satellite dual polarization SAR remote sensing images. In this study, firstly, a sliding window is used to segment the image into multiple sub-images of the same size; secondly, the iterative threshold method is used to obtain the green tide and seawater samples that have the theoretical bimodal structure from each sub-image; then, using the histogram bimodal method, the detection threshold is calculated automatically; and finally, using the threshold segmentation, the green tide areas are extracted from the images. In order to verify the proposed method, the multi-scale segmentation method and the proposed method are both used to detect green tide in China Yellow Sea. The study results show that in the case of GF-3 SAR FSII images, the proposed automatic detection method is superior to the multi-scale segmentation method, as it not only improves the accuracy of green tide detection, but also realizes the automation of green tide extraction. Furthermore, Cross-polarization (HV) images may be more suitable than co-polarization (HH) images for the extraction of green tide due to their lower noise level.

Wei, Y.; Zhang, Z.; Mu, B.; Li, Y.; Wang, Q., and Liu, R., 2020. Geolocation accuracy evaluation of GF-4 geostationary high-resolution optical images over coastal zones and offshore areas. In: Jung, H.-S.; Lee, S.; Ryu, J.H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 326-333. Coconut Creek (Florida), ISSN 0749-0208.

GF-4 is the first high-resolution geostationary satellite launched by China in 2015 for disaster prevention and mitigation. It can provide high spatial (up to 50 m) and temporal resolution optical images. Autonomous geolocation accuracy is key for the application of satellite images, particularly for coastal zones, offshore areas, and even oceans with few ground control points (GCPs). Hence, this study characterizes the autonomous geometric accuracy of GF-4 panchromatic and multispectral sensor images over coastal zones and offshore areas, which is of particular importance for the monitoring of moving targets. The evaluation of nine GF-4 images based on 630 GCPs in coastal zones and offshore areas of the Bohai Sea and Yellow Sea indicates a geolocation uncertainty of 1925 ± 976 m (39 ± 20 pixels). For GF-4 images with similar imaging geometry, the geolocation discrepancy is lower than that with different imaging geometry, with the geolocation discrepancy of 1 ± 1 and 30 ± 13 pixels, respectively. Although with the lower geolocation discrepancy, noticeable uncertainty occurs in the estimated velocity of moving targets for images with similar imaging geometry, particularly for low-speed targets, such as macro-algae. This suggests that geometric correction is needed before the speed of the moving target is extracted. The geolocation accuracy of GF-4 images through rational polynomial coefficient orthorectification without GCPs is also poor, with the geolocation uncertainty of 1350 ± 864 m, whereas the geolocation accuracy significantly improved by 8.2 times after five GCPs were added (distributed in the four corners and center of the image).

Liao, Y.; Liao, Y-Q.; Zhao, W-J.; Chen Q-H.; Li, T., and Yang, T.J., 2020. Study on mangrove of maximum likelihood: Reclassification method in Xiezhou bay. In: Jung, H.-S.; Lee, S.; Ryu, J.-H., and Cui, T. (eds.), Advances in Geospatial Research of Coastal Environments. Journal of Coastal Research, Special Issue No. 102, pp. 334-343. Coconut Creek (Florida), ISSN 0749-0208.

Xiezhou Bay in Huidong was selected as the research area to explore the mechanism of maximum likelihood value-reclassification in this paper it deepened the degree of fragmentation and plaque extraction of maximum likelihood value and establish a point-to-point valuation extraction and classification model. This paper used the maximum likelihood to extract the GF-2 image mangrove forest in Xiezhou bay, in order to establish maximum likelihood point-to-point extraction mechanism. The maximum likelihood - reclassification model can be assignment and remove plaque. By comparing the KNN classification method test and field selecting sample validation samples, the maximum likelihood value - reclassification accuracy was as high as 89.29 %. This method greatly improves the accuracy of map extraction and field conditions. At the same time, our study provided scientific support for managing and mastering mangrove technologies.