Tidal flats are associated with complicated depositional and ecological environments, and have changed considerably as a result of the erosion and sedimentation caused by tidal energy; consequently, the surface sediment distribution in tidal flats must be constantly monitored and mapped. Although several studies have been conducted with the aim of classifying intertidal surface sediments using various remote sensing methods combined with field survey, most of these studies were unable to consider various sediment types, due to the low spatial resolution of remotely sensed data. Therefore, previous studies were unable to efficiently describe precise surface sediment distribution maps. In the present study, unmanned aerial vehicle (UAV) red, green, blue (RGB) orthoimagery was used in combination with a field survey (232 samples) to produce a large-scale classification map for surface sediment distribution, in accordance with sedimentology standards, using an object-based method. The object-based method is an effective technique that can classify surface sediment distribution by analyzing its correlations with spectral reflectance, grain size, and tidal channels. Therefore, we distinguished six sediment types based on their spectral reflectance and sediment properties, such as grain composition and statistical parameters. The accuracy assessment of the surface sediment classification based on these six types indicated an overall accuracy of 72.8%, with a kappa coefficient of 0.62 and 5-m error range related to the Global Positioning System (GPS) device. We found that 11 samples were misclassified due to the effects of sun glint and cloud caused by the UAV system and shellfish beds, while 14 misclassified samples were influenced by surface water related to the elevation, tidal channels, and sediment properties. These results indicate that large-scale classification of surface sediment with high accuracy is possible using UAV RGB orthoimagery.

The first geostationary ocean color data from the Geostationary Ocean Color Imager (GOCI) onboard the Communication, Ocean, and Meteorological Satellite (COMS) have been accumulating for more than ten years from 2010. This study performs a multi-year quality assessment of GOCI chlorophyll-a (Chl-a) and radiometric data for 2012–2021 with an advanced atmospheric correction technique and a regionally specialized Chl-a algorithm. We examine the consistency and stability of GOCI, Moderate Resolution Imaging Spectroradiometer (MODIS), and Visible Infrared Imaging Radiometer Suite (VIIRS) level 2 products in terms of annual and seasonal climatology, two-dimensional frequency distribution, and multi-year time series. Overall, the GOCI agrees well with MODIS and VIIRS on annual and seasonal variability in Chl-a, as the central biological pattern of the most transparent waters over the western North Pacific, productive waters over the East Sea, and turbid waters over the Yellow Sea are reasonably represented. Overall, an excellent agreement is remarkable for western North Pacific oligotrophic waters (with a correlation higher than 0.91 for Chl-a and 0.96 for band-ratio). However, the sporadic springtime overestimation of MODIS Chl-a values compared with others is notable over the Yellow Sea and East Sea due to the underestimation of MODIS blue-green band ratios for moderate-high aerosol optical depth. The persistent underestimation of VIIRS Chl-a values compared with GOCI and MODIS occurs due to inherent sensor calibration differences. In addition, the artificially increasing trends in GOCI Chl-a (+0.48 mg m⁻³ per 9 years) arise by the decreasing trends in the band ratios. However, decreasing Chl-a trends in MODIS and VIIRS (−0.09 and −0.08 mg m⁻³, respectively) are reasonable in response to increasing sea surface temperature. The results indicate GOCI sensor degradation in the late mission period. The long-term application of the GOCI data should be done with a caveat, however; planned adjustments to GOCI calibration (2022) in the following GOCI-II satellite will essentially eliminate the bias in Chl-a trends.

Red tides caused by Margalefidinium polykrikoides occur continuously along the southern coast of Korea, where there are many aquaculture cages, and therefore, prompt monitoring of bloom water is required to prevent considerable damage. Satellite-based ocean-color sensors are widely used for detecting red tide blooms, but their low spatial resolution restricts coastal observations. Contrarily, terrestrial sensors with a high spatial resolution are good candidate sensors, despite the lack of spectral resolution and bands for red tide detection. In this study, we developed a U-Net deep learning model for detecting M. polykrikoides blooms along the southern coast of Korea from PlanetScope imagery with a high spatial resolution of 3 m. The U-Net model was trained with four different datasets that were constructed with randomly or non-randomly chosen patches consisting of different ratios of red tide and non-red tide pixels. The qualitative and quantitative assessments of the conventional red tide index (RTI) and four U-Net models suggest that the U-Net model, which was trained with a dataset of non-randomly chosen patches including non-red tide patches, outperformed RTI in terms of sensitivity, precision, and F-measure level, accounting for an increase of 19.84%, 44.84%, and 28.52%, respectively. The M. polykrikoides map derived from U-Net provides the most reasonable red tide patterns in all water areas. Combining high spatial resolution images and deep learning approaches represents a good solution for the monitoring of red tides over coastal regions.

Recent advances in deep learning (DL) and unmanned aerial vehicle (UAV) technologies have made it possible to monitor salt marshes more efficiently and precisely. However, studies have rarely compared the classification performance of DL with the pixel-based method for coastal wetland monitoring using UAV data. In particular, many studies have been conducted at the landscape level; however, little is known about the performance of species discrimination in very small patches and in mixed vegetation. We constructed a dataset based on UAV-RGB data and compared the performance of pixel-based and DL methods for five scenarios (combinations of annotation type and patch size) in the classification of salt marsh vegetation. Maximum likelihood, a pixel-based classification method, showed the lowest overall accuracy of 73%, whereas the U-Net classification method achieved over 90% accuracy in all classification scenarios. As expected, in a comparison of pixel-based and DL methods, the DL approach achieved the most accurate classification results. Unexpectedly, there was no significant difference in overall accuracy between the two annotation types and labeling data sizes in this study. However, when comparing the classification results in detail, we confirmed that polygon-type annotation was more effective for mixed-vegetation classification than the bounding-box type. Moreover, the smaller size of labeling data was more effective for detecting small vegetation patches. Our results suggest that a combination of UAV-RGB data and DL can facilitate the accurate mapping of coastal salt marsh vegetation at the local scale.