본 연구에서는 순환신경망을 이용한 댐 유입량 예측모형의 적용성 검토를 목적으로 하고 있으며, 이를 위해 소양강댐 유역 및 충주댐 유역을 대상 으로 그간 댐 운영을 통해 축적된 기상 및 수문 빅데이터를 활용하여 인공신경망 모형과 엘만 순환신경망 모형을 구축하였다. 모형의 학습과 예측 을 위하여 유역별 유입량, 강우량, 기온, 일조시간, 풍속자료가 입력자료로 사용되었고 10일간 일별 댐유입량 자료가 모델의 출력자료로 구조화 하여 학습을 진행한 후 검증을 목적으로 2016년 7월 ~ 2018년 6월까지 2개년에 대한 댐 유입량 예측을 수행하였다. 학습된 모형의 유입량 예측 결과를 비교분석한 결과, 소양강댐 유역에서는 인공신경망 모형과 순환신경망 모형 간 예측성능은 큰 차이를 보이지 않았으며, 충주댐 유역에서는 순환신경망 모형의 예측 결과가 인공신경망 모형에 비해 비교적 우수한 성능을 보임에 따라 엘만 순환신경망을 이용하여 댐 유입량 예측모형을 구축 할 경우 예측성능은 기존의 인공신경망 모형과 비슷하거나 다소 우수할 것으로 판단된다. 또한 엘만 순환신경망은 갈수기 댐 유입량 예측에 있어서 인공신경망에 비해 예측결과의 재현성이 우수한 것으로 나타났으며, 엘만 순환신경망 학습에 있어 다중 은닉층 구조가 단일 은닉층 구조보다 예측 성능 향상에 효과적인 것으로 분석되었다.

The correlation between meteorological data collected at the optical wide-field patrol network (OWL-Net) Station No. 1 and the seeing of satellite optical observation data was analyzed. Meteorological data and satellite optical observation data from June 2014 to November 2015 were analyzed. The analyzed meteorological data were the outdoor air temperature, relative humidity, wind speed, and cloud index data, and the analyzed satellite optical observation data were the seeing full-width at half-maximum (FWHM) data. The annual meteorological pattern for Mongolia was analyzed by collecting meteorological data over four seasons, with data collection beginning after the installation and initial set-up of the OWL-Net Station No. 1 in Mongolia. A comparison of the meteorological data and the seeing of the satellite optical observation data showed that the seeing degrades as the wind strength increases and as the cloud cover decreases. This finding is explained by the bias effect, which is caused by the fact that the number of images taken on the less cloudy days was relatively small. The seeing FWHM showed no clear correlation with either temperature or relative humidity.

Inactive space objects are usually rotating and tumbling as a result of internal or external forces. KOREASAT 1 has been inactive since 2005, and its drift trajectory has been monitored with the optical wide-field patrol network (OWL-Net). However, a quantitative analysis of KOREASAT 1 in regard to the attitude evolution has never been performed. Here, two optical tracking systems were used to acquire raw measurements to analyze the rotation period of two inactive satellites. During the optical campaign in 2013, KOREASAT 1 was observed by a 0.6 m class optical telescope operated by the Korea Astronomy and Space Science Institute (KASI). The rotation period of KOREASAT 1 was analyzed with the light curves from the photometry results. The rotation periods of the low Earth orbit (LEO) satellite ASTRO-H after break-up were detected by OWL-Net on April 7, 2016. We analyzed the magnitude variation of each satellite by differential photometry and made comparisons with the star catalog. The illumination effect caused by the phase angle between the Sun and the target satellite was corrected with the system tool kit (STK) and two line element (TLE) technique. Finally, we determined the rotation period of two inactive satellites on LEO and geostationary Earth orbit (GEO) with light curves from the photometry. The main rotation periods were determined to be 5.2 sec for ASTRO-H and 74 sec for KOREASAT 1.

Several optical monitoring strategies by a ground-based telescope to protect a Geostationary Earth Orbit (GEO) satellite from collisions with close approaching objects were investigated. Geostationary Transfer Orbit (GTO) objects, Inclined GeoSynchronous Orbit (IGSO) objects, and drifted GEO objects forced by natural perturbations are hazardous to operational GEO satellites regarding issues related to close approaches. The status of these objects was analyzed on the basis of their orbital characteristics in Two-Line Element (TLE) data from the Joint Space Operation Center (JSpOC). We confirmed the conjunction probability with all catalogued objects for the domestic operational GEO satellite, Communication, Ocean and Meteorological Satellite (COMS) using the Conjunction Analysis Tools by Analytical Graphics, Inc (AGI). The longitudinal drift rates of GeoSynchronous Orbit (GSO) objects were calculated, with an analytic method and they were confirmed using the Systems Tool Kit by AGI. The required monitoring area was determined from the expected drift duration and inclination of the simulated target. The optical monitoring strategy for the target area was analyzed through the orbit determination accuracy. For this purpose, the close approach of Russian satellite Raduga 1-7 to Korean COMS in 2011 was selected.

As a governmentally approved domestic entity for Space Situational Awareness, Korea Astronomy and Space Science Institute (KASI) is developing and operating an optical telescopes system, Optical Wide-field PatroL (OWL) Network. During the test phase of this system, it is necessary to determine the range of brightness of the observable satellites. We have defined standard magnitude for Low Earth Orbit (LEO) satellites to calibrate their luminosity in terms of standard parameters such as distance, phase angle, and angular rate. In this work, we report the optical brightness range of five LEO Satellites using OWL-Net.

An algorithm to automatically extract coordinate and time information from optical observation data of geostationary orbit satellites (GEO satellites) or geosynchronous orbit satellites (GOS satellites) is developed. The optical wide-field patrol system is capable of automatic observation using a pre-arranged schedule. Therefore, if this type of automatic analysis algorithm is available, daily unmanned monitoring of GEO satellites can be possible. For data acquisition for development, the COMS1 satellite was observed with 1-s exposure time and 1-m interval. The images were grouped and processed in terms of “action”, and each action was composed of six or nine successive images. First, a reference image with the best quality in one action was selected. Next, the rest of the images in the action were geometrically transformed to fit in the horizontal coordinate system (expressed in azimuthal angle and elevation) of the reference image. Then, these images were median-combined to retain only the possible non-moving GEO candidates. By reverting the coordinate transformation of the positions of these GEO satellite candidates, the final coordinates could be calculated.

Soybean (Glycine max (L.) Merr) is a short day plant and has been adapted to various climates and environments during cultivation. However, the cultivation area is restricted to a very narrow range of latitudes. To date, nine major genes (E1 to E8 and J) have been reported to control the flowering time and maturity. Here, we evaluated the role of E2, E3, E4, and their paralogue genes in late flowering soybean cultivars under long day (LD) conditions using Soybean yellow common mosaic virus (SYCMV)-based virus-induced gene silencing (VIGS) system. A total of nine VIGS constructs were infiltrated into two fully expanded cotyledons and primary leaves. After inoculation with these VIGS constructs on Jangyeobkong, which is a late-flowering cultivar, phenotypic traits were evaluated for the first flowering dates (FFDs) and pod maturities under LD conditions. The FFDs of the silenced plants occurred 50-56 days after sowing (das), while the non-silenced plants bloomed on 60-61 days. We found that the E3 paralogue-silenced plants flowered the fastest and responsive genes were identified to be associated with the promotion of flowering time. As the knock-down of E3 paralogue, expression of E1 was up-regulated, E2 was no difference, E3 and E4 genes were down-regulated in the silenced plants. Expression of GmFT2a and GmFT5a is known to be controlled by E3 and E4. Interestingly, GmFT5a were highly expressed in SYCMV:E3 paralogue-silenced plants, whereas the expression of GmFT2a was not significant. These results support that GmFT5a is able to independently promote flowering under LD conditions.