• Spatial Information Research
• /
• v.10 no.4
• /
• pp.533-550
• /
• 2002

Soil organic matter (SOM) content and other physical soil properties were extracted from a digital soil database, the Soil Survey Geographic (SSURGO) database, to map the amount of SOM and determine its relationship with topographic positions in floodplain areas along a river basin in Douglas County, Kansas. In the floodplains, results showed that slope and SOM content had a significant negative relationship. Soils near river channels were deep and nearly level, and they had the greatest SOM content in the floodplain areas. For the whole county, SOM content was influenced primarily by soil depth and percent SOM by weight. Among different slope areas, soils on mid-range slopes (10-15%) and ridgetops had the highest SOM content because they had relatively high percent SOM content by weight and very deep soils, respectively. SOM content was also significantly variable among different land cover types. Forest/woodland had significantly higher SOM content than others, followed by cropland, grassland, and urban areas.

• Korean Journal of Ecology and Environment
• /
• v.39 no.1 s.115
• /
• pp.52-61
• /
• 2006

The aim of this study was to analyze the seasonal patterns of zooplankton community dynamics in the lower Nakdong River (Mulgum, RK; river kilometer; 27 km from the estuarine barrage), with a Self-Organizing Map (SOM) based on weekly sampled data collected over ten years(1994 ${\sim}$ 2003). It is well known that zooplankton groups had important role in the food web of freshwater ecosystems, however, less attention has been paid to this group compared with other community constituents. A non-linear patterning algorithm of the SOM was applied to discover the relationship among river environments and zooplankton community dynamics. Limnological variables (water temperature, dissolved oxygen, pH , Secchi transparency, turbidity, chlorophyll a, discharge, etc.) were taken into account to implement patterning seasonal changes of zooplankton community structures (consisting of rotifers, cladocerans and copepods). The trained SOM model allocated zooplankton on the map plane with limnological parameters. Three zooplankton groups had high similarities to one another in their changing seasonal patterns, Among the limnological variables, water temporature was highly related to the zooplankton community dynamics (especially for cladocerans). The SOM model illustrated the suppression of zooplankton due to the increased river discharge, particularly in summer. Chlorophyll a concentrations were separated from zooplankton data set on the map plane, which would intimate the herbivorous activity of dominant grazers. This study introduces the zooplankton dynamics associated with limnological parameters using a nonlinear method, and the information will be useful for managing the river ecosystem, with respect to the food web interactions.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2007.05a
• /
• pp.1427-1431
• /
• 2007

In this study, the new methodology which combines Kohonen self-organizing map(KSOM) neural networks model and the conventional neural networks models such as feedforward neural networks model and generalized neural networks model is introduced to forecast flood stage in Nakdong river, Republic of Korea. It is possible to train without output data in KSOM neural networks model. KSOM neural networks model is used to classify the input data before it combines with the conventional neural networks model. Four types of models such as SOM-FFNNM-BP, SOM-GRNNM-GA, FFNNM-BP, and GRNNM-GA are used to train and test performances respectively. From the statistical analysis for training and testing performances, SOM-GRNNM-GA shows the best results compared with the other models such as SOM-FFNNM-BP, FFNNM-BP, and GRNNM-GA and FFNNM-BP shows vice-versa. From this study, we can suggest the new methodology to forecast flood stage and construct flood warning system in river basin.

• 한국방재학회:학술대회논문집
• /
• 2008.02a
• /
• pp.669-673
• /
• 2008

In this study, the new methodology which combines Kohonen self-organizing map(KSOM) neural networks model and the conventional neural networks models such as feedforward neural networks model and generalized neural networks model is introduced to forecast flood stage in Nakdong river, Republic of Korea. It is possible to train without output data in KSOM neural networks model. KSOM neural networks model is used to classify the input data before it combines with the conventional neural networks model. Four types of models such as SOM-FFNNM-BP, SOM-GRNNM-GA, FFNNM-BP, and GRNNM-GA are used to train and test performances respectively. From the statistical analysis for training and testing performances, SOM-GRNNM-GA shows the best results compared with the other models such as SOM-FFNNM-BP, FFNNM-BP, and GRNNM-GA and FFNNM-BP shows vice-versa. From this study, we can suggest the new methodology to forecast flood stage and construct flood warning system in river basin.

• The Korean Journal of Zoology
• /
• v.36 no.2
• /
• pp.175-180
• /
• 1993

Eighty one specimens of nhodeus sericeus (Pallas, 1176) were collected at the upper stream of the Som River (37$^{\circ}$32' 38" N, 127$^{\circ}$58' 57" E: Maegok-ri, Konggun-myon, Hoengsong-gun, Kanguvon-do, Korea) which is a tributary of the Namhan River. This is the first report of the species in the southern part of the Korean Peninsula. This species is well discriminated from the other species in the genus nhodeus by the following respects. The number of scales of transverse series is more than 34 (range 34-37, mean 35.7 $\pm$0.7). Head is echo감or and snout is Bonder than those of the other species. The upper part of the eve is always black and the lateral body surfaces are reddish purple. It had been hem that the southern limit of the distribution of Rhodeus sericeus was the Yonghung River in Hamgvongnam-do, north-eastern part of the Korean Peninsula. But their distributional range is extended more southerly to the Han River system which tons westward through the central part of the Korean Peninsula.Peninsula.

• Journal of Environmental Science International
• /
• v.25 no.3
• /
• pp.361-371
• /
• 2016

Ministry of Environment has been operating water quality monitoring network in order to obtain the basic data for the water environment policies and comprehensively understand the water quality status of public water bodies such as rivers and lakes. The observed water quality data is very important to analyze by applying statistical methods because there are seasonal fluctuations. Typically, monthly water quality data has to analyze that the transition comprise a periodicity since the change has the periodicity according to the change of seasons. In this study, trends, SOM and RDA analysis were performed at the Mulgeum station using water quality data for temperature, BOD, COD, pH, SS, T-N, T-P, Chl-a and Colon-bacterium observed from 1989 to 2013 in the Nakdong River. As a result of trends, SOM and RDA, the Mulgeum station was found that the water quality is improved, but caution is required in order to ensure safe water supply because concentrations in water quality were higher in the early spring(1~3 month) the most.

• Journal of Environmental Science International
• /
• v.12 no.2
• /
• pp.111-117
• /
• 2003

This paper investigates preliminarily spatial distribution soil organic matter (SOM), nitrogen (N) and phosphorus (P) and its environmental influence in wetland soil of different vegetation landscape in the Yellow River Mouth. The result shows the SOM and total nitrogen (TN), efficient N, efficient P in top layer soils of different vegetation district have significantly different content, The SOM is shown as Calamagrostis epigeios wetlands > Phragmites cmmunis wetlands > Tamarix chinensi wetlands above tidal > Suaeda salsa wetlands in high tidal > Tamarix chinensi wetlands in high tidal > tidal flats, the arrange of the TN and efficient N content is the same except that the content in Suaeda salsa wetlands in high tidal is heavier than Tamarix chinensi wetlands in high tidal. In different vegetation landscape wetland types the vertical change of soil nutrients are obvious except for p, gradually decrease from the upper to the lower. This case reflects the function of the vegetation on the wetland development of soil and proves the wetland soil has the characteristic of new born and bad degree of development. SOM, TN, efficient N and efficient P content in wetland soils have significantly positive correlation, but TP have no correlation with them but efficient p. The contents of TN in wetland soils range from 58~1480 mg/kg, total average content 408 mg/kg, average content of above 30 cm is 625 mg/kg. The range of TP content in the soil is 372~1042 mg/kg, total average is 569 mg/kg, average content of above 20 cm is 611 mg/kg. According the P it occurs mainly as calcium phosphates, and the validity is lower, therefore, N and P in the new born wetlands cannot produce serious impact on the environments at present.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.249-249
• /
• 2017

본 연구에서는 낙동강 본류의 진동과 적포교, 남강의 거룡강, 대산, 정암 수위표의 관측된 2015년부터 2016년 2년간 수위자료를 이용하여, Self-Organizing Maps(SOM)과 LOcally WEighted Scatter plot Smoother(Lowess) 기법으로 패턴과 추세를 분석하였다. SOM 분석 결과, 낙동강 본 류의 진동과 적포교, 남강의 거룡강, 대산은 동일한 패턴과 추세를 나타냈다. 수위의 범위도, SOM 분석에서 진동 최소 EL. 4.41m, 최대 EL. 5.01m 범위, 적포교 최소 EL. 4.56m, 최대 EL. 5.38m 범위, 거룡강 최소 EL. 4.53m, 최대 EL. 5.18m 범위, 대산 최소 EL. 4.57m, 최대 EL. 5.35m 범위로 큰 차이가 발생하지 않았다. 거룡강과 대산 수위 관측지점은 낙동강 본류의 배수위 영향을 받는 것을 알 수 있었으며, 두 지점의 수위 관측 목적에 따라 상류로 지점 변경이 필요 할 수도 있을 것으로 판단된다.

• Journal of Korean Society on Water Environment
• /
• v.33 no.1
• /
• pp.78-89
• /
• 2017

Water quality monitoring network data is being affected continuously due to non-point source pollution arising from agricultural land located on the Gwangsancheon outlet in the Nakdong River basin. In this study, we have performed analysis of water quality monitoring system, water quality pattern using SOM and water quality in the Gwangsancheon for sub-basin located at Gisan-myeon in the Nakdong River basin. We have developed and applied the model to estimate the runoff and non-point source loading. As a result of SOM pattern, the effect of non-point source pollution was the largest in the paddy fields and fields. As a result of the developed model, we found contribution rate and reduction rate for non-point source loading according to change of landuse because the reduction effect of nonpoint pollutants was 20.9% of SS, 9.9% of TN, 21.2% of TP and 8.9% of TOC depending on the landuse change.

• Journal of Ecology and Environment
• /
• v.39 no.1
• /
• pp.61-70
• /
• 2016

Major objectives of the study were to analyze chemical and biological influences of the river ecosystem on the artificial weir construction at three regions of Sejong-Weir (S_j-W), Gongju-Weir (G_j-W), and Baekje-Weir (B_j-W) during 2008-2012. After the weir construction, the discharge volume increased up to 2.9 times, and biological oxygen demand (BOD) and electrical conductivity (EC) significantly decreased (p < 0.05). Also, the decrease of total phosphorus (TP) was also evident after the weir construction, but still hyper-eutrophic conditions, based on criteria by , were maintained. Multi-metric model of Index of Biological Integrity (IBI) showed that IBI values averaged 21.0 (range: 20-22; fair condition) in the B_wc, and 14.3 (range: 12-18; poor condition) in the A_wc. The model values of IBI in S_j-W and G_j-W were significantly decreased after the weir construction. The model of Self-Organizing Map (SOM) showed that two groups (cluster I and cluster II) of B_wc and A_wc were divided in the analysis based on the clustering map trained by the SOM. Principal Component Analysis (PCA) was similar to the results of the SOM analysis. Taken together, this research suggests that the weir construction on the river modified the discharge volume and the physical habitat structures along with distinct changes of some chemical water quality. These physical and chemical factors influenced the ecosystem health, measured as a model value of IBI.