• Proceedings of the Korea Water Resources Association Conference
• /
• 2020.06a
• /
• pp.291-291
• /
• 2020

제주도는 강수의 지표침투성이 좋은 화산섬의 지질특성상 지표수의 개발이용여건이 취약한 관계로 용수의 대부분을 지하수에 의존하고 있다. 따라서 제주도는 정책 및 연구적으로 오랜 기간동안 지하수의 보전관리에 많은 노력을 기울여 오고 있다. 하지만 최근 기후변화로 인한 강수의 변동성 증가로 인해 지하수위의 변동성 또한 증가할 가능성이 있으며 따라서 지하수위의 급격한 하강에 대비하여 지하수위의 예측 및 지하수 취수량 관리의 필요성이 요구되고 있다. 지하수에 절대적으로 의존하고 있는 제주도의 수자원 이용 여건을 고려할 때, 지하수의 취수량 관리를 위한 지하수위의 실시간 예측이 필요한 실정이다. 하지만 기존의 예측방법에 의한 제주도 지하수위 예측기간은 충분히 길지 않으며 예측기간이 길어지면 예측성능이 낮아지는 문제점이 있었다. 본 연구에서는 이러한 단점을 보완하기 위해 딥러닝 알고리즘인 Long Short Term Memory(LSTM)를 활용하여 제주도 남동쪽 표선유역 중산간지역의 1개 지하수위 관측정에 대해 지하수위를 예측하고 분석하였다. R 기반의 Keras 패키지에 있는 LSTM 알고리즘을 사용하였고, 입력자료는 인근의 성판악 및 교래 강우관측소의 일단위 강수량자료와 인근 취수정의 지하수 취수량자료 및 연구대상 관측정의 지하수위 자료를 사용하였으며, 사용된 자료의 기간은 2001년 2월 11일부터 2019년 10월 31일까지 이다. 2001년부터 13년의 보정 및 3년의 검증용 시계열자료를 사용하여 매개변수의 보정 및 과적합을 방지하였고, 3년의 예측용 시계열자료를 사용하여 LSTM 알고리즘의 예측성능을 평가하였다. 목표 예측일수는 1일, 10일, 20일, 30일로 설정하였으며 보정, 검증 및 예측기간에 대한 모의결과의 평가지수로는 Nash-Sutcliffe Efficiency(NSE)를 활용하였다. 모의결과, 보정, 검증 및 예측기간에 대한 1일 예측의 NSE는 각각 0.997, 0.997, 0.993 이었고, 10일 예측의 NSE는 각각 0.993, 0.912, 0.930 이었다. 20일 예측의 경우 NSE는 각각 0.809, 0.781, 0.809 이었으며 30일 예측의 경우 각각 0.677, 0.622, 0.633 이었다. 이것은 LSTM 알고리즘에 의한 10일 예측까지는 관측 지하수위 시계열자료를 매우 적절히 모의할 수 있다는 것을 의미하며, 20일 예측 또한 적절히 모의할 수 있다는 것을 의미한다. 따라서 LSTM 알고리즘을 활용하면 본 연구대상지점에 대한 2주일 또는 3주일의 안정적인 지하수위 예보가 가능하다고 판단된다. 또한 LSTM 알고리즘을 통한 실시간 지하수위 예측은 지하수 취수량 관리에 활용할 수 있을 것이다.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.199-207
• /
• 2024

Local reservoirs are crucial sources for agricultural water supply, necessitating stable water level management to prepare for extreme climate conditions such as droughts. Water level prediction is significantly influenced by local climate characteristics, such as localized rainfall, as well as seasonal factors including cropping times, making it essential to understand the correlation between input and output data as much as selecting an appropriate prediction model. In this study, extensive multivariate data from over 400 reservoirs in Jeollabuk-do from 1991 to 2022 was utilized to train and validate a water level prediction model that comprehensively reflects the complex hydrological and climatological environmental factors of each reservoir, and to analyze the impact of each input feature on the prediction performance of water levels. Instead of focusing on improvements in water level performance through neural network structures, the study adopts a basic Feedforward Neural Network composed of fully connected layers, batch normalization, dropout, and activation functions, focusing on the correlation between multivariate input data and prediction performance. Additionally, most existing studies only present short-term prediction performance on a daily basis, which is not suitable for practical environments that require medium to long-term predictions, such as 10 days or a month. Therefore, this study measured the water level prediction performance up to one month ahead through a recursive method that uses daily prediction values as the next input. The experiment identified performance changes according to the prediction period and analyzed the impact of each input feature on the overall performance based on an Ablation study.

• Journal of the Korean association of regional geographers
• /
• v.2 no.1
• /
• pp.51-67
• /
• 1996

There are three main rice-growing regions in the United States: the prairie region along the Mississippi River Valley in eastern Arkansas; the Gulf Coast prairie region in southwestern Louisiana and southeastern Texas; and the Central Valley of California. The Central Valley of California is producing about 23% of the US rice(Fig. 1). In California. most of the crop has been produced in the Colusa, Sutter, Butte, Glenn Counties of the Sacramento Valley since 1912, when rice was commercially grown for the first time in the state(Fig. 2). Roughly speaking, the average annual area sown to rice in California is about 300,000 acres to 400,000 acres during the last forty years(Fig. 3). California rice is grown under a Mediterranean climate characterized by warm, dry, clear days, and a long growing season favorable to high photosynthetic rates and high rice yields. The average rice yield per acre is probably higher in California than in any other rice-growing regions of the world(Fig. 4). A dependable supply of irrigation water must be available for a successful rice culture. Most of the irrigation water for California rice comes from the winter rain and snow-fed reservoir of the Sierra Nevada mountain ranges. Less than 10 percent of rice irrigation water is pumped from wells in areas where surface water is not sufficient. It is also essential to have good surface drainage if maximum yields are to be produced. Rice production in California is highly mechanized, requiring only about four hours of labor per acre. Mechanization of rice culture in California includes laser-leveler technology, large tractors, self-propelled combines for harvesting, and aircraft for seeding, pest control, and some fertilization. The principal varieties grown in California are medium-grain japonica types with origins from the cooler rice climates of the northern latitudes (Table 1). Long-grain varieties grown in the American South are not well adapted to California's cooler environment. Nearly all the rice grown recently in California are improved into semidwarf varieties. Choice of variety depends on environment, planting date, quality desired, marketing, and harvesting scheduling. The Rice Experiment Station at Biggs is owned, financed, and administered by the rice industry. The station was established in 1912, as a direct result of the foresight and effort of Charles Edward Chambliss of the United States Department of Agriculture. Now, The station's major effort is the development of improved rice varieties for California.

• Journal of The Korean Society of Grassland and Forage Science
• /
• v.34 no.1
• /
• pp.52-59
• /
• 2014

This study was conducted to estimate the Hanwoo carrying capacity when whole crop wheat, as a winter forage crop, was grown on mixed-sowing of legume forage and by applying cattle manure on the productivity and feed value of whole crop wheat during the period of 2012~2013. The experiment was conducted in a split plot design with three replications. The main plots consisted of three different culture methods such as whole crop wheat and mixed sowing combination with hairy vetch or forage pea. The subplots consisted of four different applications of cattle manure (0, 50, 100 and 150 kg N/ha). The annual amount of dry matter (DM) of whole crop wheat in Gyeongju were higher than those of in Gyeongsan and Yeongju, and the mixed-sowing of hairy vetch mixture was the highest (p<0.05) compared with the single-sowing of whole crop wheat and mixed-sowing of legume in Gyeongju. The DM amounts were increased proportionately corresponded to the applying level of cattle manure, and was found to be significantly (p<0.05) high at the level of 100 and 150 kg/ha groups. With the feed value of forage, the crude protein (CP) contents tended to be higher in the mixed-sowing of legume than the single-sowing of whole crop wheat at the wintering experimental sites of legume. For whole crop wheat, total digestible nutrients (TDN) content was the highest in the mixed sowing plots of forage pea in Gyeongsan; however, there was no significant differences among the single-sowing of whole crop wheat. The carrying capacity of Hanwoo (head/ha) was higher (p<0.05) in Gyeongju (3.83 head) than that in Gyeongsan (3.11) and Yeongju (1.35). Further, the carrying capacity in the single-sowing of whole crop wheat was lower than that in the mixed-sowing of legume, and the hairy vetch was the highest among the mixed-sowing groups (p<0.05). Overall, the present results recommend taking into account the wintering for the mix-sowing of legume in Gyeongbuk province. The cattle manure may be applied for legume in the wintering unavailable regions. In addition, applying cattle manure at the level of 100~150 kg/ha and the mix-sowing of legume may increase the productivity per unit area and feed value, including the CP, for improving the carrying capacity of Hanwoo.

• Korean Journal of Soil Science and Fertilizer
• /
• v.42 no.6
• /
• pp.439-446
• /
• 2009

Nitrogen balance in the regional scale which was calculated the difference between nitrogen input and output was estimated to assess the impact of rice cultivation on water environment. Nitrogen balances in Gyeonggi province, where nitrogen concentration in irrigation water was high and in Chungnam province, where nitrogen absorbtion by rice was high, were -5.4 and -8.3 kg $-8.3kg\;ha^{-1}\;yr^{-1}$, respectively. Nitrogen balances of paddy field in Gangwon province, where nitrogen output was small and irrigation water was clean, and in Gyeongnam province, where organic matter content of soil was high and rice yield was low, were 4.9 and $14.0kg\;ha^{-1}\;yr^{-1}$, respectively. Average nitrogen balance and total nitrogen absorption of paddy field in Korea were estimated to $-0.3kg\;ha^{-1}\;yr^{-1}$ and $-3,315Mg\;yr^{-1}$, respectively. When the nitrogen concentration in irrigation water was increased by $1mg \;L^{-1}$, nitrogen balance of rice paddy changed by $-2.91kg\;ha^{-1}\;yr^{-1}$. Also, when nitrogen fertilizer applied was decreased from 110 to $90kg\;ha^{-1}$ and the same harvest was maintained, the nitrogen absorption by rice paddy from irrigation water was estimated to increase by 10,600 Mg per year in Korea. However, in cases, the harvest was reduced to either 90% or 85%, nitrogen balances were changed from -11.7 to -2.3 and $2.4kg\;ha^{-1}$, respectively. These results suggest that the reduction of nitrogen fertilizer use may not always lead to a negative nitrogen balance and sustainable agriculture can achieve by not cutting down the use of fertilizer only but by reduction of fertilizer application concurrently by maintenance of harvest and by utilization of environmental characteristics such as nutrient contents in irrigation water and soils.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.54 no.4
• /
• pp.127-135
• /
• 2012

This paper introduced the flow forecast modeling system that a water management agency in west central Florida, Tampa Bay Water has been operated to forecast monthly rainfall and streamflow in the Tampa Bay region, Florida. We evaluated current 1-year monthly rainfall forecasts and flow forecasts and actual observations to investigate the benefits of incorporating rainfall forecasts into monthly flow forecast. Results for rainfall forecasts showed that the observed annual cycle of monthly rainfall was accurately reproduced by the $50^{th}$ percentile of forecasts. While observed monthly rainfall was within the $25^{th}$ and $75^{th}$ percentile of forecasts for most months, several outliers were found during the dry months especially in the dry year of 2007. The flow forecast results for the three streamflow stations (HRD, MB, and BS) indicated that while the 90 % confidence interval mostly covers the observed monthly streamflow, the $50^{th}$ percentile forecast generally overestimated observed streamflow. Especially for HRD station, observed streamflow was reproduced within $5^{th}$ and $25^{th}$ percentile of forecasts while monthly rainfall observations closely followed the $50^{th}$ percentile of rainfall forecasts. This was due to the historical variability at the station was significantly high and it resulted in a wide range of forecasts. Additionally, it was found that the forecasts for each station tend to converge after several months as the influence of the initial condition diminished. The forecast period to converge to simulation bounds was estimated by comparing the forecast results for 2006 and 2007. We found that initial conditions have influence on forecasts during the first 4-6 months, indicating that FMS forecasts should be updated at least every 4-6 months. That is, knowledge of initial condition (i.e., monthly flow observation in the last-recent month) provided no foreknowledge of the flows after 4-6 months of simulation. Based on the experimental flow forecasts using the observed rainfall data, we found that the 90 % confidence interval band for flow predictions was significantly reduced for all stations. This result evidently shows that accurate short-term rainfall forecasts could reduce the range of streamflow forecasts and improve forecast skill compared to employing the stochastic rainfall forecasts. We expect that the framework employed in this study using available observations could be used to investigate the applicability of existing hydrological and water management modeling system for use of stateof-the-art climate forecasts.

• Korean Journal of Environmental Agriculture
• /
• v.26 no.3
• /
• pp.233-238
• /
• 2007

To assess pollutant loads in Nakdong river from highland agriculture in Kyungbuk province we. analyzed water qualities such as BOD, COD, T-N, T-P and SS in year 2005. BOD values in rainy period (June and July) were relatively higher than those in dry period, and those in 4 sites among 17 sites ranged from 10.71-19.25 mg/L which exceeded water criteria (8 mg/L) for agricultural use. COD values showed similar trends like BOD values. These trends might be caused by outflow of nutrients applied in agricultural lands. T-N content ranged from 0.1 to 14 mg/L. Those in lower reaches of stream were greater in those in upper stream. Compared to T-N contents during non-farming season, T-N content in farming season were higher. These phenomenon could be due to continuous input of nutrients from small watercourses. Averaged T-P content in lower stream during farming season was 0.4 mg/L, which was eight times higher than the limiting level for algae occurrence (0.05 mg/L). BOD, T-P, T-N loads from alpine agricultural practices were 12.25 $kg/km^2{\cdot}day$, 0.55 $kg/km^2{\cdot}day$ and 32.35 $kg/km^2{\cdot}day$, respectively. These values were greater than those from forestry. Therefore, Best management Practices (BMP) for alpine agricultural field are needed to reduce pollutant loads in Nakdong river.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.22 no.3
• /
• pp.10-20
• /
• 2019

Irregular rainfall caused by climate change, in combination with non-point pollution, can cause water systems worldwide to suffer from frequent eutrophication and algal blooms. This type of water pollution is more common in agricultural prone to water system inflow of non-point pollution. Therefore, in this study, the correlation between Unmanned Aerial Vehicle(UAV) multi-spectral images and total phosphorus, total nitrogen, and chlorophyll-a with indirect association of algal blooms, was analyzed to identify the usability of UAV image to identify water quality characteristics in agricultural streams. The analysis the vegetation index Normalized Differences Index (NDVI), the Normalized Differences Red Edge(NDRE), and the Chlorophyll Index Red Edge(CIRE) for the detection of multi-spectral images and algal blooms collected from the target regions Yang cheon and Hamyang Wicheon. The analysis of the correlation between image values and water quality analysis values for the water sampling points, total phosphorus at a significance level of 0.05 was correlated with the CIRE(0.66), and chlorophyll-a showed correlation with Blue(-0.67), Green(-0.66), NDVI(0.75), NDRE (0.67), CIRE(0.74). Total nitrogen was correlated with the Red(-0.64), Red edge (-0.64) and Near-Infrared Ray(NIR)(-0.72) wavelength at the significance level of 0.05. The results of this study confirmed a significant correlations between multi-spectral images collected through UAV and the factors responsible for water pollution, In the case of the vegetation index used for the detection of algal bloom, the possibility of identification of not only chlorophyll-a but also total phosphorus was confirmed. This data will be used as a meaningful data for counterplan such as selecting non-point pollution apprehensive area in agricultural area.

• Journal of the Korean Geographical Society
• /
• v.30 no.3
• /
• pp.242-254
• /
• 1995

The Hadano Basin is located at a distance of about 70kms and 60kms from Tokyo and Yokohama and lies in the south-west part of the Kanto region in Japan. The basin area, which correspoends to the catchment of the Kaname River, is about areal size of 60.7$\textrm{km}^2$ and extends about length of 8kms in E-W direction and about width of 5kms in N-S direction (Fig.1). The Hadano basin is filled with thick pile of the alluvum from deposits composed of volcanic materials, mostly came from the Hakone Volcano and overlain by Fuji Volcanic ashes. Fluvial deposits form the good aquifer, therefore water resources of Handano City has been largely depending upon the eroundwater. Urbanization and industrialization of the basin has been rapid in the last thirty years, after activation of "Factory Attraction Policy of Hadano City" in 1956. Growth in population and number of factory due to urbanization changed the land-use pattern of the basin rapidly and increased the water demands. Therefore, Hadano City exploited a new source of water supply, and have introduced the prefectureal waterworks since 1976. On the other hand, the rapid urbanization has brought about the pollution of streams in the basin by domestic sewage and industrial waste water. Diffusion rate of sewerage systems in Hadano City is 38% in 1993. In ordcr to examine the impact of anthropogenic factors on river environments, the author took up the change of land-use and diffusion area of sewerage as parameters, and performed field surveys on water discharge and quality. The survey has been made at upstream and downstream of the main stream regularly per month, to get informati ons about the variation of discharge and water quality aiong the stream and its diurnal fluctuation. Annual variation has been analyzed based the data from Hadano City Office. The results are summarized as follows. 1. Stream discharge has been increasing by urbanization (Fig.3). Water quality (C $l^{-10}$ , N $H^{+}$$_{ 4}$-N, BOD) has been improving gradually after the application of sewerage service, yet water pollution load at the lower station has increased than that at the upper one because of the larger anthropogenic discharge volumes (Fig.4). 2. Corrclation coefficient of discharges between upper and lower was 0.81-0.92. Pollutant loads of the R. Kamame after the confluence with R. Kuzuha grew up by 2.4-3.7 times as compared with its upper reaches, and it increased to 3.7-6.9 times after the confluence with the R. Muro (Fig.5). 3. The changes of water quality along the stream can be divided into two groups (Fig.6a). First: water quality of the R. Kaname and R. Shijuhachisse is becoming worse towards the lower reaches because the water from branches are polluted. Second: water quality are improved in the lower where spring and small branch streams supply clear water, for example R. Mizunashi, R. Muro and R. Kuzuha. 4. Measured discharge at the upper station in the R. Shijuhachisse is 0.153㎥/sec, and about 55% of this is recharged until it reaches to the lower point. The R. Mizunashi has a discharge of 1.155㎥/sec at the upper point, is recharged 0.24㎥/sec until the midstream and groundwater spring 0.2㎥/sec at the lower reaches. R. Kuzuha recharged all the mountain runoff (0.2㎥/sec) at the upper reaches. The R. Muro is supplied by many springs and the estimated discharge of spring was 0.47㎥/sec (Fig.6b). 5. Diurmal variations in discharge and water quality are influenced clearly by domestic and industrial waste waters (Fig.7, 8).ed clearly by domestic and industrial waste waters (Fig.7, 8).

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.12 no.2
• /
• pp.87-94
• /
• 1979

In order to study the water pollution in Suyeong Bay, Busan, some chemical constituents were determined at 25 stations in the neap tides on 9 Aug. 1977 and spring tides on 30 Aug. 1977. Range and mean values of the constituents in the spring tides are as follows: $pH\;6.54\~8.06,\;7.54;$ electrical conductivity $0.413\~0.481\times10^5\;\mu\mho/cm,\;0.467\times10^5\;\mu\mho/cm;\;transparency\;0.2\~5.5m,\;2.2m;$ turbidity $1\~60ppm$, 14ppm, chlorosity $15.20\~18.11g/\ell,\;17.67g/\ell;$ fluoride ion $0.94\~1.03ppm$, 0.99ppm; dissolved oxygen $0.17\~7.60ppm$, 4.77ppm; sulfide $0\~0.46ppm$, 0.07ppm; chemical oxygen demand $1.20\~40.74ppm$, 6.11ppm; ammonia-nitrogen $0.060\~0.520ppm$, 0.180ppm; nitrite-nitrogen $0.001\~0.026ppm$, 0.009ppm; nitrate-nitrogen $0\~0.037ppm$, 0.014ppm; phosphate-phosphorus $0.002\~0.261ppm$, 0.050ppm; n-Hexane soluble $0.5\~5.4ppm$, 2.1ppm ; iron $1.0\~104.11\;ppb$, 24.15ppb ; copper $0\~27.45ppb$, 4.19ppb; lead $0\~2.50ppb$, 0.92ppb; zinc $0\~5.15ppb$, 1.47ppb ; cadmium $0\~0.26ppb$, 0.04ppb; and mercury $0.05\~0.37ppb$, 0.11ppb respectively. The variations of the contents of the chemical constituents in the spring tides were larger than in the neap tides. The contents of COD, sulfide, nutrient salts and heavy metals were the highest in the estuary of Suyeong River, and decreased in order of off Kwangan-Ri region, outer Bay and off Haeun-Dae region. The water quality in Suyeong Bay was particularly shown that the concentrations of COO, iron, copper and mercury were higher than those of other coastal aseas and deficiency in dissolved oxygen was observed in some parte of Suyeong Bay. In consideration of the relationship between the chlorosity and the concentrations of nutrient salts, COD and total heavy metals, water pollution of this area is considered due to the inflow of Suyeong River which was extremely polluted by sewage and industrial wastewaters.