• Journal of Korean Society of Rural Planning
• /
• v.30 no.1
• /
• pp.57-66
• /
• 2024

This study aimed to develop a precise vegetation cover classification model for small streams using the combination of drone remote sensing and support vector machine (SVM) techniques. The chosen study area was the Idong stream, nestled within Geosan-gun, Chunbuk, South Korea. The initial stage involved image acquisition through a fixed-wing drone named ebee. This drone carried two sensors: the S.O.D.A visible camera for capturing detailed visuals and the Sequoia+ multispectral sensor for gathering rich spectral data. The survey meticulously captured the stream's features on August 18, 2023. Leveraging the multispectral images, a range of vegetation indices were calculated. These included the widely used normalized difference vegetation index (NDVI), the soil-adjusted vegetation index (SAVI) that factors in soil background, and the normalized difference water index (NDWI) for identifying water bodies. The third stage saw the development of an SVM model based on the calculated vegetation indices. The RBF kernel was chosen as the SVM algorithm, and optimal values for the cost (C) and gamma hyperparameters were determined. The results are as follows: (a) High-Resolution Imaging: The drone-based image acquisition delivered results, providing high-resolution images (1 cm/pixel) of the Idong stream. These detailed visuals effectively captured the stream's morphology, including its width, variations in the streambed, and the intricate vegetation cover patterns adorning the stream banks and bed. (b) Vegetation Insights through Indices: The calculated vegetation indices revealed distinct spatial patterns in vegetation cover and moisture content. NDVI emerged as the strongest indicator of vegetation cover, while SAVI and NDWI provided insights into moisture variations. (c) Accurate Classification with SVM: The SVM model, fueled by the combination of NDVI, SAVI, and NDWI, achieved an outstanding accuracy of 0.903, which was calculated based on the confusion matrix. This performance translated to precise classification of vegetation, soil, and water within the stream area. The study's findings demonstrate the effectiveness of drone remote sensing and SVM techniques in developing accurate vegetation cover classification models for small streams. These models hold immense potential for various applications, including stream monitoring, informed management practices, and effective stream restoration efforts. By incorporating images and additional details about the specific drone and sensors technology, we can gain a deeper understanding of small streams and develop effective strategies for stream protection and management.

• Journal of Wetlands Research
• /
• v.11 no.3
• /
• pp.21-35
• /
• 2009

To propose a sound direction for river rehabilitation plan, the naturalness of the river must be evaluated. In this study, the conventional methods for evaluating the naturalness of rivers proposed by Park, et al. has been improved to estimate the weights for each items subject to assessment and applied adequate Analytic Hierarchy Process (AHP). To estimate the weights, survey was conducted on a group of 83 experts in Civil and Environmental Engineering with a customized questionnaire. The results of the survey were generalized and used to estimate the weights of each items. Among the 14 evaluation items that have been generated as a result of the estimation, a total of 7 weight values such as water quality, variations in water flow, vegetation around minor bed, curve of waterway, vegetation around flood plain, variety in minor bed width, and diversity of stream bed materials appeared to be high. The improved method for evaluating the naturalness of rivers was applied to Hoe-ya River, where the method has successfully discriminated the sections having low degree of naturalness. The proposed assessment method can be employed as a criterion to determine the business zones in the river environment maintenance projects.

• Korean Journal of Ecology and Environment
• /
• v.55 no.1
• /
• pp.19-34
• /
• 2022

This study was identified the distribution of vegetation in domestic estuaries and analysed the relationship with environmental fcators based on the health assessment data of the estuary ecosystem surveyed between 2016 and 2018. Of the 325 estuaries surveyed, 187 vegetation communites were investigated in 300 estuaries and 53 halophyte communites accounted for 28.3%. No vegetation distribution was found in the other 25 estuaries. Considering the circulation of estuary, 41 halophyte communites were investigated in open estuaries and 26 halophyte communites in closed estuaries. As a result of canonical correlation analysis (CCA) between the entire distributed vegetation community and environmental factors, salinity (conductivity), T-N, and T-P concentrations were strongly correlated. Among the riverbed material materials of the estuary, the composition ratio of silt, sand, and pebble was the most correlated. Therefor, it is estimated that the distribution location of the halophyte communites were differentiated by being influenced by the sea ares, estuary circulation type, resistance to salinity and riverbed material ratio.

• Ecology and Resilient Infrastructure
• /
• v.1 no.2
• /
• pp.68-81
• /
• 2014

In this study, the numerical simulation regarding the process and characteristics of topography change due to the vegetation recruitment and growth was carried out by adding the vegetation growth model to two-dimensional flow and sediment transport models. The vegetation introduction and recruitment on the condition for developing an alternate bar reduced the bar migration. The vegetated area and channel width changes were more significantly influenced by changes in upstream discharge rather than the duration of low flow. When the upstream discharge decreased, the vegetation area increased and the channel width decreased. The vegetation introduction and recruitment on the condition for developing a braided channel significantly influenced the characteristics of topography changes. In the braided channel, vegetation reduced the braided index, and when the upstream discharge decreased significantly, the channel topography was changed from the braided channel to the single channel. The vegetation area decreased as the upstream discharge increased. The channel width decreased significantly after the vegetation was introduced and it also decreased as the upstream discharge decreased. It was confirmed through the numerical simulation that a decrease in flood discharge accelerated the vegetation introduction and recruitment in the channel and this allowed to confirm its influence on the characteristics of topography changes qualitatively.

• 한국생태학회:학술대회논문집
• /
• 2002.08a
• /
• pp.113-118
• /
• 2002

Vegetation and riverine structures were compared among urban and several semi-natural rivers by applying direct gradient analysis and ordination method. Urban rivers showed different species composition from the semi-natural ones. Species composition of semi-natural rivers depended on the geographical positions, such as the upstream, midstream, and downstream and on micro-topographies, such as the waterside, flooding bed, and embankment. Semi-natural rivers showed gentle change in micro-topography, whereas urban one did not so. Our restoration plan to improve the ecological quality of the degraded urban river by imitating semi-natural river was prepared based on those data.

• The Korean Journal of Ecology
• /
• v.25 no.4
• /
• pp.247-252
• /
• 2002

Vegetation and riverine structures were compared among urban and several semi-natural rivers by applying direct gradient analysis and ordination method. Urban rivers showed different species composition from the semi-natural ones. Species composition of semi-natural rivers depended on the geographical positions, such as the upstream, midstream, and downstream and on micro-topographies, such as the waterside, flooding bed, and embankment. Semi-natural rivers showed gentle change in micro-topography, whereas urban one did not so. Our restoration plan to improve the ecological quality of the degraded urban river by imitating semi-natural river was prepared based on those data.

• ALGAE
• /
• v.23 no.1
• /
• pp.83-90
• /
• 2008

Eelgrass beds are very productive and provide nursery functions for a variety of fish and shellfish species. Management for the conservation of eelgrass beds along the Korean coasts is critical, and requires comprehensive strategies such as vegetation mapping. We suggest a mapping method to spatial distribution and quantify of eelgrass beds using a digital echosounder. Echosounding data were collected from the northeast part of Kwangyang Bay, on the south of Korea, in March, 2007. A transducer was attached to a boat equipped with a DGPS. The boat completed a transect survey scanning whole eelgrass beds of 11.7 km2 with a speed of 1.5-2 m s-1 (3-4 knot). The acoustic reflectivity of eelgrass allowed for detection and explicit measurements of canopy cover and height. The results showed that eelgrass bed was distributed in depth from 1.19 to 3.6 m (below MSL) and total dry weight biomass of 4.1 ton with a vegetation area of 4.05 km2. This technique was found to be an effective way to undertake the patch size and biomass of eelgrass over large areas as nondestructive sampling.

• Journal of the Korean Society of Environmental Restoration Technology
• /
• v.10 no.3
• /
• pp.71-87
• /
• 2007

In order to improve and manage rationally for Youngdam creek in Haman-gun, the channel change, water quality and ecological adaptation etc. were evaluated by monitoring during three years. The distinct differences of channel characteristics appeared in the extents of bed excavation and the revetment methods because overall reconstruction was occurred intense disturbance of width expansion and levee construction etc. The scour and deposition were caused according to channel characteristics of ripple and pool, and the adaptation of vegetation was distinguished from each revetment methods clearly. Water quality was maintained first class on the basis of BOD as 0.5~1.5mg/${\ell}$. The vegetation appeared in 380 species close-to-nature river improvement before, but 64 species it after one year, 159 species it after two year and 158 species it after three year. The animal life appeared in 123 species of close-to-nature river improvement before, but 103 species it after one year, 116 species it after two year and 119 species it after three year.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.39 no.2
• /
• pp.307-316
• /
• 2019

Development of vegetation in stream channel increases resistance to flow, resulting in increase in river stage upon flood and affecting change in stage-discharge relationship. Vegetation revealed in stream by water level reaching a peak and then declined upon flood is mostly found as prone. Taking an account of flow distribution with the number of vegetation, prone vegetation layer might be at height where discharge rate is zero (0) (Stephan and Guthnecht, 2002). However, there is a tendency that flow rate is overestimated when applying the height of river bed to flow area with no consideration of the height of vegetation layer in flow rate by float measurement. In this study, reliable flow measurement in stream with vegetation was calculated by measuring the height of vegetation layer after flood and excluding the vegetation layer-projected area from the flow area. The result showed the minimum 4.34 % to maximum 10.82 % of flow deviation depending on the scale of discharge. Accordingly, reliable velocity-area methods would be determined if vegetation layer-projected area in stream is considered in flow rate estimation using the flow area during the flood.

• Ecology and Resilient Infrastructure
• /
• v.8 no.4
• /
• pp.212-222
• /
• 2021

In this study, the flow resistance coefficient was calculated considering the physical properties and distribution characteristics of floodplain vegetation, and the effect of floodplain vegetation distribution on flow characteristics was analyzed by reflecting it in a two-dimensional numerical simulation. The three-dimensional point clouds of vegetation acquired using ground lidar were analyzed to apply floodplain vegetation's physical properties to the existing formula for vegetation flow resistance calculation. The floodplain vegetation distribution in the modeling was divided into locally distributed and fully distributed conditions in the floodplain. As a result of the simulation of the study site, the flow resistance coefficient of floodplain vegetation was found to have a value of about five times or more compared to the flow resistance coefficient of the main channel bed when the design flood occurs based on Manning's n coefficient. Also, it affected the hydraulic characteristics in the main channel and floodplain.