• 천문학회지
• /
• 제40권3호
• /
• pp.67-82
• /
• 2007

The basic building block of solar filaments/prominences is thin threads of cool plasma. We have studied the spectral properties of velocity threads, clusters of thinner density threads moving together, by analyzing a sequence of $H{\alpha}$ images of a quiescent filament. The images were taken at Big Bear Solar Observatory with the Lyot filter being successively tuned to wavelengths of -0.6, -0.3, 0.0, +0.3, and +0.6 ${\AA}$ from the centerline. The spectra of contrast constructed from the image data at each spatial point were analyzed using cloud models with a single velocity component, or three velocity components. As a result, we have identified a couple of velocity threads that are characterized by a narrow Doppler width($\Delta\lambda_D=0.27{\AA}$), a moderate value of optical thickness at the $H{\alpha}$ absorption peak($\tau_0=0.3$), and a spatial width(FWHM) of about 1". It has also been inferred that there exist 4-6 velocity threads along the line of sight at each spatial resolution element inside the filament. In about half of the threads, matter moves fast with a line-of-sight speed of $15{\pm}3km\;s^{-1}$, but in the other half it is either at rest or slowly moving with a line-of-sight velocity of $0{\pm}3km\;s^{-1}$. It is found that a statistical balance approximately holds between the numbers of blue-shifted threads and red-shifted threads, and any imbalance between the two numbers is responsible for the non-zero line-of-sight velocity determined using a single-component model fit. Our results support the existence not only of high speed counter-streaming flows, but also of a significant amount of cool matter either being at rest or moving slowly inside the filament.

• 한국지리정보학회지
• /
• 제20권4호
• /
• pp.65-76
• /
• 2017

본 연구는 범죄와 환경과의 관계가 밀접한 영향을 미친다는 전제하에 범죄억제를 위한 정책수립 및 설계를 위한 대안마련의 사전단계로 범죄위험지역 분석과 요인, 공간적 특성이 반영된 예측모델을 제시하는 것을 주요 목적으로 수행하였다. 연구의 분석결과, H구 전체 지역에 범죄가 균등하게 분포하였을 경우와 대비하여 약 1/4 수준으로 범죄발생지역은 군집성을 나타내고 있었으며 주로 H 해수욕장 배후 상업지역에서 특정범죄(강간, 절도, 폭력)는 강한 핫스팟을 나타내었다. 결과적으로 독립변수와 종속변수의 공간적 상관관계를 고려하는 지리가중회귀모형을 이용함으로써 보다 효과적으로 적용할 수 있음을 알 수 있었다. 이러한 결과는 다중선형회귀 모형의 결과에서와 같이 연구지역 내 폐공가수가 종속변수와 단순히 부적의 관계를 가지는 것이 아니라, 지역에 따라 종속변수에 서로 다른 영향을 미치는 것을 알 수 있다.

• 한국음향학회지
• /
• 제34권4호
• /
• pp.303-309
• /
• 2015

화자 분할은 사전에 분류되지 않은 데이터를 각각의 화자로 분류하는 연구이며 DBSCAN(Density-Based Spatial Clustering of Applications with Noise)은 간결함과 계산의 효율성으로 인해 화자분할 분야에 널리 사용되어 왔다. 그러나 클러스터의 데이터들이 공간적이지 않으며 서로 다른 클러스터가 근접하여 경계를 공유할 때 오버클러스터링 문제가 발생하여 DBSCAN의 성능이 하락한다. 본 논문에서는 DBSCAN과 문제점을 설명하고, 개체의 지역적 특성에 기반한 밀도 기반 클러스터링 알고리즘을 제안한다. 제안하는 알고리즘은 개체의 지역적 밀도와 분산의 정도에 따라 가변적인 판단 기준을 탐색에 이용한다. DBSCAN과 제안 기법의 실험을 통해 성능을 비교하고 제안 기법의 효용을 보인다. 실험 결과 제안한 방법은 오버클러스터링이 발생하지 않으며 DBSCAN에 비해 보다 높은 정확도를 보여 지역적 특성을 이용한 접근 방법이 효과적임을 증명한다.

• 한국경제지리학회지
• /
• 제23권4호
• /
• pp.408-424
• /
• 2020

본 연구는 커피전문점 소재지별 지가 속성을 토대로 생존하여 영업 중인 커피전문점과 폐업한 커피전문점을 대상으로 공간 군집화를 시도하고, 군집 유형별 토지 특성(지가 및 필지 정보), 교통 요인(물리적 접근성)과 점포 속성(프랜차이즈 정보, 소재지별 동종업종의 개·폐업 경험), 공간 밀집도(커널 밀도 추정) 등 다양한 입지 특성을 통해 커피전문점의 생멸 분포에서 나타나는 특징을 파악한다. 이를 위해 영업 중 및 폐업한 커피전문점의 공간 군집은 일반적 입지분포형, 주거지역의 상업화형, 상업중심지 입지형 등으로 유형 구분하고, 군집 유형별 입지 특성을 비교 분석한다. 군집 결과, 영업 중 및 폐업한 커피전문점의 입지는 같은 유형으로 분류되더라도 서로 다른 공간 분포 양상을 나타내고, 입지 유형이 달라도 특정 핫스폿에서는 모두 높은 밀집도를 나타낸다. 본 연구의 분석 결과는 커피전문점 창업을 비롯해 지역별 상권정보를 파악하는데 기초자료로 제공될 수 있다.

• 천문학회보
• /
• 제40권1호
• /
• pp.57.3-57.3
• /
• 2015

We present a test observation result of the KMTNet Intensive Nearby Southern Galaxy group Survey (KINGS). The KINGS is designed to study nearby galaxy groups (NGC 55, NGC 253, NGC 5128, and M83 groups), taking the advantage of the wide field coverage of the KMTNet. The main goal of the KINGS is to produce extensive catalogs of dwarf galaxies, ultra compact dwarfs (UCDs), and intraglobular clusters in the galaxy groups. We will also investigate the spatial distribution of intragroup light in each group. We present a progress report of the project based on the test BVI observations of two galaxy groups. We discuss the result from the test observation and possible improvement for future observations.

• 천문학회지
• /
• 제37권5호
• /
• pp.563-570
• /
• 2004

Turbulence is a crucial component of dynamics of astrophysical fluids dynamics, including those of ISM, clusters of galaxies and circumstellar regions. Doppler shifted spectral lines provide a unique source of information on turbulent velocities. We discuss Velocity-Channel Analysis (VCA) and its offspring Velocity Coordinate Spectrum (VCS) that are based on the analytical description of the spectral line statistics. Those techniques are well suited for studies of supersonic turbulence. We stress that a great advantage of VCS is that it does not necessary require good spatial resolution. Addressing the studies of mildly supersonic and subsonic turbulence we discuss the criterion that allows to determine whether Velocity Centroids are dominated by density or velocity. We briefly discuss ways of going beyond power spectra by using of higher order correlations as well as genus analysis. We outline the relation between Spectral Correlation Functions and the statistics available through VCA and VCS.

• 천문학회지
• /
• 제38권2호
• /
• pp.23-31
• /
• 2005

CN and CH band strengths for fourteen bright giants in the globular cluster M71 have been measured from archival spectra obtained with the Multiple Mirror Telescope. Adding the collected. data from the literature we confirm a bimodality of CN distribution on the red giant branch and the honzontal branch, and CN-CH anti-correlations on the lower giant branch and horizontal branch. However a CN-CH anti-correlation on the upper red giant branch is not quite clear as those of other branches. The small number If statistics could not be excluded as a possible cause. To confirm this, a greater number of sample stars are needed. We also confirm that the ratio of CN-strong to CN-weak stars is quite different from that in 47 Tuc, although the anti-correlation between CN and CH bands, the bimodality of the CN distribution, and the spatial distribution of CN stars in M71 are found to be similar to those III 47 Tuc.

• 천문학회지
• /
• 제35권1호
• /
• pp.29-34
• /
• 2002

We have conducted UBVI CCD photometry of an intermediate-age open cluster NGC 559 to investigate the effect of dynamical evolution on the stellar distributions in NGC 559. Our photometry allows better estimates of distance and age of the cluster owing to much deeper photometry (V $\le$ 21) than previous ones. It is found that the luminosity function and mass function as well as the spatial stellar distributions are affected by the dynamical evolution. Mass segregation leads to the central concentration of the high mass stars, which results in the flattened mass function inside the half mass radius.

• 천문학회지
• /
• 제43권4호
• /
• pp.115-121
• /
• 2010

UBVI CCD photometry is obtained for the Ruprecht 93 (Ru 93)region. We are unable to confirm the existence of an intermediate-age open cluster in Ru 93 from the spatial distribution of blue stars. On the other hand, we find two young star groups in the observed field: the nearer one (Ru 93 group) comprises the field young stars in the Sgr-Car arm at $d{\approx}2.1$ kpc, while the farther one (WR 37 group) is the young stars around WR 37 at $d{\approx}4.8$ kpc. We derive an abnormal extinction law ($R_V$ = 3.5) in the Ruprecht 93 region.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2004년도 Proceedings of ISRS 2004
• /
• pp.428-431
• /
• 2004

A multi-stage hierarchical clustering technique, which is an unsupervised technique, has been proposed in this paper for classifying the hyperspectral data .. The multistage algorithm consists of two stages. The 'local' segmentor of the first stage performs region-growing segmentation by employing the hierarchical clustering procedure with the restriction that pixels in a cluster must be spatially contiguous. The 'global' segmentor of the second stage, which has not spatial constraints for merging, clusters the segments resulting from the previous stage, using a context-free similarity measure. This study applied the multistage hierarchical clustering method to the data generated by band reduction, band selection and data compression. The classification results were compared with them using full bands.