• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.164-164
• /
• 2017

토지이용 및 토지피복 변화는 인간의 활동에 따른 결과를 반영하며 자연환경의 변화와도 밀접한 관련이 있다. 수문환경의 변화는 수로 건설이나 하천정비, 사진개간, 도시화 등과 같은 인위적인 변화와 지구온난화에 따라 변화하는 기후 패턴에도 많은 영향을 받고 있다. 본 연구에서는 안성천 공도수위관측소 상류유역을 대상으로 SWAT(Soil and Water Assessment) 모형을 이용하여 현재의 유역 수문환경조건을 보다 현실적으로 분석하기 위해 연구를 실시하였다. 유역의 수문환경 조건을 정량적으로 파악하기 위하여 CLUE-s 모형을 이용한 토지이용변화가 기후변화 시나리오에 따른 수문학적 거동(지표유출, 증발산량, 총유출량, 수질거동)에 미치는 영향을 SWAT모형을 이용하여 파악하고자 하였다. 안성천의 공도 수위관측소($366.5km^2$)을 대상으로 유역 내 3지점의 기상관측소(이천, 수원, 천안)를 대상으로 40년(1976~2015)동안의 일 기상자료를 수집하여 SWAT 모형을 구축하였다. 평년 및 가뭄년을 포함하는 선별된 기간에 대하여 다양한 목적함수($R^2$, NSE, RMSE)를 활용하여 일반적인 전체기간 검보정 방법이 아닌 각각의 조합된 기간의 극단적 유출특성에 초점을 맞추어 검보정을 수행하였다. 이와 같이, 앞서 실시한 연구로 모형의 매개변수들을 해당 유역에 가뭄과 같은 극단적 유출상황에 따라 선행보정 실시하였으며, 이는 이번 연구에서의 모형의 신뢰성 있는 수문 및 수질 거동 분석을 할 수 있을 것이라 생각된다. 선행 연구 자료를 이용하여 토지이용변화를 고려한 본 연구를 통해 미래의 유역 수문환경조건 변화에 따라 수자원을 정량적으로 파악하는 데에 도움이 될 수 있을 것으로 판단된다.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.63 no.6
• /
• pp.101-115
• /
• 2021

The purpose of this study is to evaluate the compatibility of the vegetation index between the two satellites and the applicability of agricultural monitoring by comparing and verifying NDVI (Normalized Difference Vegetation Index) based on Sentinel-2 and Terra MODIS (Moderate Resolution Imaging Spectroradiometer). Terra MODIS NDVI utilized 16-day MOD13Q1 data with 250 m spatial resolution, and Sentinel-2 NDVI utilized 10-day Level-2A BOA (Bottom Of Atmosphere) data with 10 m spatial resolution. To compare both NDVI, Sentinel-2 NDVIs were reproduced at 16-day intervals using the MVC (Maximum Value Composite) technique. As a result of time series NDVIs based on two satellites for 2019 and compare by land cover, the average R² (Coefficient of determination) and RMSE (Root Mean Square Error) of the entire land cover were 0.86 and 0.11, which indicates that Sentinel-2 NDVI and MODIS NDVI had a high correlation. MODIS NDVI is overestimated than Sentinel-2 NDVI for all land cover due to coarse spatial resolution. The high-resolution Sentinel-2 NDVI was found to reflect the characteristics of each land cover better than the MODIS NDVI because it has a higher discrimination ability for subdivided land cover and land cover with a small area range.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.5
• /
• pp.1-10
• /
• 2019

Due to the climate change the drought had been occurring more frequently in recent two decades as compared to the previous years. The change in the pattern and frequency of the rainfall have a direct effect on the farming sector; therefore, the quantitative estimation of water supply is necessary for efficient agricultural water reservoir management. In past researches, there had been several studies conducted in estimation and evaluation of water supply based on the irrigational water requirement. However, some researches had shown significant differences between the theoretical and observed data based on this requirement. Thus, this study aims to propose an approach in estimating reservoir rate based on empirical method that utilized observed reservoir rate data. The result of these two methods in comparison with the previous one is seen to be more fitted for both R2 and RMSE with the observed reservoir rate. Among these procedures, the method that considers the drought year data shows more fitted outcomes. In addition, this new method was verified using 15-year (2002 to 2006) linear regression equation and then compare the preceeding 3-year (1999 to 2001) data to the theoretical method. The result using linear regression equation is also perceived to be more closely fitted to the observed reservoir rate data than the one based on theoretical irrigation water requirement. The new method developed in this research can therefore be used to provide more suitable supply data, and can contribute to effectively managing the reservoir operation in the country.

• Journal of Climate Change Research
• /
• v.8 no.2
• /
• pp.125-143
• /
• 2017

This study is conducted to estimate potential evapotranspiration of 10 weather observing systems in Andong Dam watershed with FAO56 Penman-Monteith (FAO56 PM) methodology using the meteorological data from 2013 to 2014. Also, assuming that there is no solar radiation data, humidity data or wind speed data, the potential evapotranspiration was estimated by FAO56 PM and the results were evaluated to discuss whether the methodology is applicable when meteorological dataset is not available. Then, the potential evapotranspiration was estimated with Hargreaves method and compared with the potential evapotranspiration estimated by FAO56 PM only with the temperature dataset. As to compare the potential evapotranspiration estimated from the complete meteorological dataset and that estimated from limited dataset, statistical analysis was performed using the Root Mean Square Error (RMSE), the Mean Bias Error (MBE), the Mean Absolute Error (MAE) and the coefficient of determination ($R^2$). Also the Inverse Distance Weighted (IDW) method was performed to conduct spatial analysis. From the result, even when the meteorological data is limited, FAO56 PM showed relatively high accuracy in calculating potential evapotranspiration by estimating the meteorological data.

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.5
• /
• pp.89-99
• /
• 2019

In this study, we estimated the spatially-distributed soil moisture at the high resolution ($10m{\times}10m$) using the satellite-based Sentinel-1A/B SAR (Synthetic Aperture Radar) sensor images. The Sentinel-1A/B raw data were pre-processed using the SNAP (Sentinel Application Platform) tool provided from ESA (European Space Agency), and then the pre-processed data were converted to the backscatter coefficients. The regression equations were derived based on the relationships between the TDR (Time Domain Reflectometry)-based soil moisture measurements and the converted backscatter coefficients. The TDR measurements from the 51 RDA (Rural Development Administration) monitoring sites were used to derive the regression equations. Then, the soil moisture values were estimated using the derived regression equations with the input data of Sentinel-1A/B based backscatter coefficients. Overall, the soil moisture estimates showed the linear trends compared to the TDR measurements with the high Pearson's correlations (more than 0.7). The Sentinel-1A/B based soil moisture values matched well with the TDR measurements with various land surface conditions (bare soil, crop, forest, and urban), especially for bare soil (R: 0.885~0.910 and RMSE: 3.162~4.609). However, the Mandae-ri (forest) and Taean-eup (urban) sites showed the negative correlations with the TDR measurements. These uncertainties might be due to limitations of soil surface penetration depths of SAR sensors and complicated land surface conditions (artificial constructions near the TDR site) at urban regions. These results may infer that qualities of Sentinel-1A/B based soil moisture products are dependent on land surface conditions. Although uncertainties exist, the Sentinel-1A/B based high-resolution soil moisture products could be useful in various areas (hydrology, agriculture, drought, flood, wild fire, etc.).

• Journal of The Korean Society of Agricultural Engineers
• /
• v.61 no.1
• /
• pp.1-8
• /
• 2019

Developing reliable soil moisture prediction techniques at agricultural regions is a pivotal issue for sustaining stable crop productions. In this study, a physically-based SWAP(Soil-Water-Atmosphere-Plant) model was suggested to estimate soil moisture dynamics at the study sites. ROSETTA was also integrated to derive the soil hydraulic properties(${\alpha}$, n, ${\Theta}_r$, ${\Theta}_s$, $K_s$) as the input variables to SWAP based on the soil information(Sand, Silt and Clay-SSC, %). In order to predict the soil moisture dynamics in future, the mid-term TIGGIE(THORPEX Interactive Grand Global Ensemble) and long-term S2S(Subseasonal to Seasonal) weather forecasts were used, respectively. Our proposed approach was tested at the six study sites of RDA(Rural Development Administration). The estimated soil moisture values based on the SWAP model matched the measured data with the statistics of Root Mean Square Error(RMSE: 0.034~0.069) and Temporal Correlation Coefficient(TCC: 0.735~0.869) for validation. When we predicted the mid-/long-term soil moisture values using the TIGGE(0~15 days)/S2S(16~46 days) weather forecasts, the soil moisture estimates showed less variations during the TIGGE period while uncertainties were increased for the S2S period. Although uncertainties were relatively increased based on the increased leading time of S2S compared to those of TIGGE, these results supported the potential use of TIGGE/S2S forecasts in evaluating agricultural drought. Our proposed approach can be useful for efficient water resources management plans in hydrology, agriculture, etc.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.215-215
• /
• 2017

NRCS-CN 방법은 총 강우량으로부터 유출량을 계산하는 방법으로, 국내에서는 설계홍수량 산정 시 NRCS-CN 방법의 사용을 권장하고 있다. CN값은 토지이용 및 피복, 토양특성, 수문학적 조건(AMC)에 따른 함수로 결정할 수 있으나, 보통의 경우 미국의 National Engineering Handbook (NEH-4)에서 제시한 표를 활용한다. 그러나, 우리나라의 토지피복 및 토지이용 현황은 미국과 다르기 때문에 현실 조건을 반영한 조정이 필요함에도 불구하고, 충분한 관측 자료가 확보되지 않아 이러한 조정이 어려운 실정이다. NRCS-CN 방법에서는 결과 값이 총 강수량보다 CN에 크게 의존적이기 때문에 부정확한 CN 값의 산정은 큰 오차를 야기할 수 있다. 또한 소유역에서는 초기손실량이 설계홍수량 산정에 큰 영향을 미치지만 우리나라는 초기손실률을 20%의 고정된 값을 일괄적으로 적용하고 있으며, 이는 제주도와 같은 특수한 투수성 지층에서는 적합하지 않다는 지적을 받아왔다. 여러 선행연구에서 강수량과 CN 사이에는 특정 관계식이 존재하며, 고정된 CN 값이 아닌 강수량에 따라 변화하는 값을 적용하는 것이 기존의 NRCS-CN 방법보다 더 정확한 결과를 나타낸다는 것이 확인된 바 있다. 본 연구에서는 NRCS-CN 방법의 CN 값과 초기손실률을 유역에 적합하게 개선하기 위해서 기존의 NRCS-CN 모형에 점근 유출곡선지수방법(Asymptotic CN Regression Method)을 통해 산정된 CN값과 각기 다른 초기손실률(0.01, 0.05, 0.10, 0.20, 0.40)을 적용하여 개선된 총 8개의 모형을 한강 권역 소유역에 적용하였다. RMSE, MAE 및 R-square 등의 지표를 이용하여 모형 검정을 수행하였으며, 최적의 모형 및 미개변수를 선정하였다. 그 결과 기존의 NRCS-CN 방법보다 점근 유출곡선지수방법을 적용했을 때 더 작은 오차를 나타내는 것을 확인하였으며, 대부분의 유역에서 0.01 또는 0.05 등 기존보다 더 작은 초기손실률을 채택 시 실측값과 가장 적은 오차를 나타냈다.

• Journal of Korean Society on Water Environment
• /
• v.35 no.5
• /
• pp.425-431
• /
• 2019

Recently, people have been attracting attention to the good quality of water resources as well as water welfare. to improve the quality of life. This study is a papers on the prediction of benthic macroinvertebrate index (BMI), which is a aquatic ecological health, using the machine learning based CFS (Correlation-based Feature Selection) method and the random forest model to compare the measured and predicted values of the BMI. The data collected from the Han River's branch for 10 years are extracted and utilized in 1312 data. Through the utilized data, Pearson correlation analysis showed a lack of correlation between single factor and BMI. The CFS method for multiple regression analysis was introduced. This study calculated 10 factors(water temperature, DO, electrical conductivity, turbidity, BOD, $NH_3-N$, T-N, $PO_4-P$, T-P, Average flow rate) that are considered to be related to the BMI. The random forest model was used based on the ten factors. In order to prove the validity of the model, $R^2$, %Difference, NSE (Nash-Sutcliffe Efficiency) and RMSE (Root Mean Square Error) were used. Each factor was 0.9438, -0.997, and 0,992, and accuracy rate was 71.6% level. As a result, These results can suggest the future direction of water resource management and Pre-review function for water ecological prediction.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.22 no.4
• /
• pp.340-345
• /
• 2020

In this study, we developed the quality control and assurance method of measurement data of SPAR (Soil-Plant-Atmosphere-Research) system, a climate change research facility, for the first time. It was found that the precise processing of CO2 flux data among many observations were sig nificantly important to increase the accuracy of canopy photosynthesis measurements in the SPAR system. The collected raw CO2 flux data should first be removed error and missing data and then replaced with estimated data according to photosynthetic lig ht response curve model. Comparing the correlation between cumulative net assimilation and soybean biomass, the quality control and assurance of the raw CO2 flux data showed an improved effect on canopy photosynthesis evaluation by increasing the coefficient of determination (R2) and lowering the root mean square error (RMSE). These data processing methods are expected to be usefully applied to the development of crop growth model using SPAR system.

• Journal of ILASS-Korea
• /
• v.27 no.2
• /
• pp.57-65
• /
• 2022

The purpose of this study is to use machine learning to build a model capable of predicting the flash boiling spray characteristics. In this study, the flash boiling spray was visualized using Shadowgraph visualization technology, and then the spray image was processed with MATLAB to obtain quantitative data of spray characteristics. The experimental conditions were used as input, and the spray characteristics were used as output to train the machine learning model. For the machine learning model, the XGB (extreme gradient boosting) algorithm was used. Finally, the performance of machine learning model was evaluated using R² and RMSE (root mean square error). In order to have enough data to train the machine learning model, this study used 12 injectors with different design parameters, and set various fuel temperatures and ambient pressures, resulting in about 12,000 data. By comparing the performance of the model with different amounts of training data, it was found that the number of training data must reach at least 7,000 before the model can show optimal performance. The model showed different prediction performances for different spray characteristics. Compared with the upstream spray angle and the downstream spray angle, the model had the best prediction performance for the spray tip penetration. In addition, the prediction performance of the model showed a relatively poor trend in the initial stage of injection and the final stage of injection. The model performance is expired to be further enhanced by optimizing the hyper-parameters input into the model.