• 한국수자원학회논문집
• /
• 제47권9호
• /
• pp.801-812
• /
• 2014

본 연구에서는 합천댐, 회천 유역에 대하여 비슬산 이중편파 레이더를 이용하여 유역평균강우량을 각각 추정하여 지상 우량계의 강우량과 비교한 후 분포형 유출모형인 Vflo 모형을 사용하여 유출을 모의하였다. 유역평균 강우량 분석 결과 이중편파 레이더 변수를 사용한 결과 첨두우량 및 전체 강우량 값이 지상우량계 값과 근접하게 산정되었고, Vflo 분포형 모형을 사용한 유출모의 분석 결과 또한 이중편파 레이더 변수를 사용한 유출모의가 실제 유출량에 근접하였다. 이는 이중편파 레이더를 사용한 강우추정 및 유출모의가 기존의 수평반사도만을 사용한 단일편파 레이더보다 정확성을 높일 수 있는 방법이라고 판단된다.

• 대기
• /
• 제24권3호
• /
• pp.365-378
• /
• 2014

The limits of S-band dual-polarization radars in Korea are not reflected on the recent weather forecasts of Korea Meteorological Administration and furthermore, they are only utilized for rainfall estimations and hydrometeor classification researches. Therefore, this study applied four merging methods [SA (Simple Average), WA (Weighted Average), SSE (Sum of Squared Error), TV (Time-varying mergence)] to the QPE (Quantitative Precipitation Estimation) model [called RAR (Radar-AWS Rainfall) calculation system] using single-polarization radars and S-band dual-polarization radar in order to improve the accuracy of the rainfall estimation of the RAR calculation system. As a result, the merging results of the WA and SSE methods, which are assigned different weights due to the accuracy of the individual model, performed better than the popular merging method, the SA (Simple Average) method. In particular, the results of TVWA (Time-Varying WA) and TVSSE (Time-Varying SSE), which were weighted differently due to the time-varying model error and standard deviation, were superior to the WA and SSE. Among of all the merging methods, the accuracy of the TVWA merging results showed the best performance. Therefore, merging the rainfalls from the RAR calculation system and S-band dual-polarization radar using the merging method proposed by this study enables to improve the accuracy of the quantitative rainfall estimation of the RAR calculation system. Moreover, this study is worthy of the fundamental research on the active utilization of dual-polarization radar for weather forecasts.

• 한국수자원학회논문집
• /
• 제48권4호
• /
• pp.257-268
• /
• 2015

본 연구에서는 이중편파 레이더 추정강우의 홍수예보 활용성을 평가하였다. 비슬산 강우레이더 100 km 반경 내 AWS (Automatic Weather System) 123개 관측소를 대상으로 레이더 추정강우의 오차를 레이더 반경 및 강우강도의 증가에 따라 평가하였다. 이중편파 레이더 추정강우가 단일편파 레이더 추정강우에 비해 오차가 작은 것으로 확인되었다. 또한, 이중편파 레이더 추정강우의 홍수예보 활용성 평가 및 적용을 위해 유역평균강우량을 산정하여 평가하였다. 평가 결과, 이중편파 레이더 추정강우가 단일편파 레이더 추정강우에 비해 관측치에 유사하게 나타났으며, 강우형태에 관계없이 강우 강도가 강한 부분에서 이중편파 레이더의 정확도가 향상됨을 보였다. 그러나 차등반사도를 통해 산정된 강우는 과대추정되는 경향이 나타났다. 연속형 저류함수모형인 SURR 모형에 적용하여 남강댐 유역에 대한 유출해석을 수행하였다. 이중편파 레이더 추정강우를 통한 유출량이 단일편파 레이더 추정강우에 비해 유출용적오차는 약 12~63%, 첨두유량오차는 약 30~42% 감소하였으며, 평균제곱근오차 또한 감소하는 것으로 나타났다. 또한 이중편파 레이더에 의해 산정된 유역평균강우량을 유출모형에 적용할 경우 AWS 강우로부터 추정된 유출결과보다 더 우수한 경우가 있어 향후 홍수예보 활용 시 예보의 정확도 향상에 기여하리라 판단된다.

• 한국수자원학회논문집
• /
• 제47권12호
• /
• pp.1227-1237
• /
• 2014

이중편파레이더는 강수의 형태를 구분하고 대기 중의 기상 현상뿐만 아니라 비강수에코에 대한 정보를 제공하기 때문에 보다 정확한 강수량 추정을 가능하게 한다. 그러나 수직, 수평으로 진동하는 전파를 송 수신하여 생성되는 이중편파레이더 관측변수들은 레이더 자체가 갖는 시스템적 관측오차를 포함하기 때문에 정량적 강수량 추정을 위해서는 이에 대한 보정이 필수적이다. 본 연구에서는 2차원 영상우적계(2-Dimensional Video Disdrometer, 이하 2DVD) 관측 자료를 이용하여 비슬산 이중편파레이더가 갖는 Z, $Z_{DR}$ 관측오차를 계산한 후, 관측오차 보정에 따라 강수량이 정량적으로 얼마나 개선되는지를 살펴보았다. 총 33강수사례에 대한 분석결과, Z는 약-0.3~5.5 dB, $Z_{DR}$는 -0.1~0.6 dB의 관측오차를 가지며, 대부분의 사례에서 Z와 $Z_{DR}$는 모의된 값보다 낮게 관측하였다. 관측오차를 보정한 전 후 산출된 이중편파레이더 강수량 추정값을 지상관측 강우강도와 비교한 결과, 평균 bias와 RMSE는 각각 1.54 mm/hr, 1.73 mm/hr로 보정 전의 1.69 mm/hr, 2.54 mm/hr 보다 감소함으로써 지상우량계 관측값 대비 레이더 강수량 추정값이 약 7~61% 향상되었다.

• 한국수자원학회:학술대회논문집
• /
• 한국수자원학회 2020년도 학술발표회
• /
• pp.166-166
• /
• 2020

Recently, As the frequency of localized heavy rains increases, the use of high-resolution radar data is increasing. The produced radar rainfall has still gaps of spatial and temporal compared to gauge observation rainfall, and in many studies, various statistical techniques are performed for correct rainfall. In this study, the precipitation correction of the S-band Dual Polarization radar in use in the flood forecast was performed using the ConvAE algorithm, one of the Convolutional Neural Network. The ConvAE model was trained based on radar data sets having a 10-min temporal resolution: radar rainfall data, gauge rainfall data for 790minutes(July 2017 in Cheongju flood event). As a result of the validation of corrected radar rainfall were reduced gaps compared to gauge rainfall and the spatial correction was also performed. Therefore, it is judged that the corrected radar rainfall using ConvAE will increase the reliability of the gridded rainfall data used in various physically-based distributed hydrodynamic models.

• 한국수자원학회논문집
• /
• 제54권1호
• /
• pp.27-38
• /
• 2021

최근 레이더 강우량을 수문학적으로 활용하기 위한 연구가 활발하게 진행되고 있다. 하지만 기상레이더의 경우 관측 특성상 산악 효과 등으로 인한 관측 영역의 한계로 빔의 차폐가 발생하며, 이는 강우량의 과소 추정의 원인이 된다. 이에 본 연구에서는 기상 레이더의 수문학적 활용을 위해 Hybrid Surface Rainfall (HSR)기법을 활용하여 레이더 강우를 추정하였으며, 분포형 유출모형인 GRM 모형을 활용하여 유출 해석을 수행하였다. 5개의 강우 사상에 대해 강우 보정 및 유출 모의를 수행한 결과 HSR 기법을 활용한 이중편파 레이더 강우량(Q_H_KDP)이 지상 강우량의 15% 이내의 오차를 보이며 정확성이 가장 높았으며, 이를 활용하여 유출 해석을 수행한 결과 역시 R2 0.9 이상 NRMSE가 0.8 이하, NSE 0.5 이상의 정확도를 나타내었다. 본 연구를 통하여 이중편파 레이더 강우의 적용성을 확인할 수 있었으며, 향후 레이더의 수문학적 활용에 관련된 연구에 활용성이 클 것으로 판단된다.

• 대기
• /
• 제29권2호
• /
• pp.197-211
• /
• 2019

The observational sensitivity of dual-polarization weather radar was quantitatively analyzed by using two different pulse widths. For this purpose, test radar scan strategy which consisted of consecutive radar scan using long (LP: $2{\mu}s$) and short (SP: $1{\mu}s$) pulses at the same elevation angle was employed. The test scan strategy was conducted at three operational S-band dual-polarization radars (KSN, JNI, and GSN) of Korea Meteorological Administration (KMA). First, the minimum detectable reflectivity (MDR) was analyzed as a function of range using large data set of reflectivity ($Z_H$) obtained from JNI and GSN radars. The MDR of LP was as much as 7~22 dB smaller than that of SP. The LP could measure $Z_H$ greater than 0 dBZ within the maximum observational range of 240 km. Secondly, polarimetric observations and the spatial extent of radar echo between two pulses were compared. The cross-polar correlation coefficient (${\rho}_{hv}$) from LP was greater than that from SP at weak reflectivity (0~20 dBZ). The ratio of $Z_H$ (> 0 dBZ) and ${\rho}_{hv}$(> 0.95) bin to total bin calculated from LP were greater than those from SP (maximum 7.1% and 13.2%). Thirdly, the frequency of $Z_H$ (FOR) during three precipitation events was analyzed. The FOR of LP was greater than that of SP, and the difference in FOR between them increased with increasing range. We conclude that the use of LP can enhance the sensitivity of polarimetric observations and is more suitable for detecting weak echoes.

• Journal of electromagnetic engineering and science
• /
• 제10권3호
• /
• pp.183-185
• /
• 2010

A 2.4 GHz dual-polarized antenna for a Doppler radar is studied. The proposed dual-polarized antenna using a stacked annular ring patch with two co-centric gap-coupled feed lines and a $90^{\circ}$ hybrid exhibits fairly good performance of 22 dB isolation at a center frequency of 2.4 GHz. Using a $90^{\circ}$ hybrid, a right-handed circular polarization for the transmitter and a left-handed circular polarization for the receiver are implemented. The gain of the designed antenna is about 0 dBi over operating frequencies. The antenna size including a ground plane is only $40{\times}40\;mm^2$.

• 대기
• /
• 제27권4호
• /
• pp.467-481
• /
• 2017

A correction method of reflectivity in partial beam blockage (PBB) area is suggested, which is based on the combination of digital terrain information and self-consistency principle between polarimetric observation. First, the reflectivity was corrected by adding the radar energy loss estimated from beam blockage simulation using digital elevation model (DEM) and beam propagation geometry in standard atmosphere. The additional energy loss by unexpected obstacles and non-standard beam propagation was estimated by using the coefficient between accumulated reflectivity ($Z_H$) and differences of differential phase shift (${\Phi}_{DP}$) along radial direction. The proposed method was applied to operational S-band dual-polarization radar at Jindo and its performance was compared with those of simulation method and self-consistency method for six rainfall cases. When the accumulated reflectivity and increment of ${\Phi}_{DP}$ along radial direction are too small, the self-consistency method has failed to correct the reflectivity while the combined method has corrected the reflectivity bias reasonably. The correction based on beam simulation showed the underestimation. For evaluation of rainfall estimation, the FBs (FRMSEs) of simulation method and self-consistency principle were -0.32 (0.59) and -0.30 (0.57), respectively. The proposed method showed the lowest FB (-0.24) and FRMSE (0.50). The FB and FMSE were improved by about 18% and by 19% in comparison to those before correction (-0.42 and 0.70). We can conclude that the proposed method can improve the accuracy of rainfall estimation in PBB area.

• 대한원격탐사학회지
• /
• 제33권1호
• /
• pp.47-60
• /
• 2017

The permafrost active layer plays an important role in permafrost dynamics. Ecological patterns, processes, and water and ice contents in the active layer are spatially and temporally complex depending on landscape heterogeneity and local-scale variations in hydrological processes. Although there has been emerging interest in the application of optical remote sensing techniques to permafrost environments, optical sensors are significantly limited in accessing information on near surface geo-cryological conditions. The primary objective of this study was to investigate capability of L-band SAR data for monitoring spatio-temporal variability of permafrost ecosystems and underlying soil conditions. This study exploits information from different polarimetric SAR observables in relation to permafrost environmental conditions. Experimental results show that each polarimetric radar observable conveys different information on permafrost environments. In the case of the dual-pol mode, the radar observables consist of two backscattering powers and one correlation coefficient between polarimetric channels. Among them, the dual-pol scattering powers are highly sensitive to freeze/thaw transition and can discriminate grasslands or ponds in thermokarst area from other permafrost ecosystems. However, it is difficult to identify the ground conditions with dual-pol observables. Additional backscattering powers and correlation coefficients obtained from quad-pol mode help understanding seasonal variations ofradar scattering and assessing geo-cryological information on soil layers. In particular, co-pol coherences atHV-basis and circular-basis were found to be very usefultools for mapping and monitoring near surface soil properties.