• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2019년도 춘계학술대회
• /
• pp.341-343
• /
• 2019

고농도 미세먼지 발생이 증가함에 따라 미세먼지 예측에 많은 관심이 집중되고 있다. 미세먼지는 대기 중에 있는 직경 $10{\mu}m$ 이하의 밀입자 물질을 말하며, 온도, 상대습도, 풍속 등의 기상 변화에 영향을 받는다. 따라서 미세먼지 예측을 위해 기상 정보와의 상관관계를 분석하는 다양한 연구가 진행되었다. 하지만 미세먼지의 비선형적 시계열 분포는 예측 모델의 복잡도를 증가시키고, 부정확한 예측값을 초래할 수 있다. 본 연구에서는 기계학습의 군집 알고리즘 및 분류알고리즘을 이용하여 미세먼지의 비선형적 특성을 완화하고자 한다. 사용된 기계학습 알고리즘은 병합군집, 밀도기반군집이며, 각 알고리즘을 통한 군집결과를 비교, 분석하였다.

• 한국연초학회지
• /
• 제21권1호
• /
• pp.102-108
• /
• 1999

This study was conducted to evaluate the ventilation to remove gases, vapor and particles of environmental tobacco smoke(ETS) in a closed room. The ventilation rate choosed were 0.445 ㎥/min, 0.528 ㎥/min, and 0.625 ㎥/min. ETS components measured were total suspended particle(TSP), ultraviolet particulate matter(UVPM), fluorescent particulate matter(FPM), solanesol, carbon dioxide($CO_2$), carbon monoxide(CO), nicotine, and 3-ethenylpyri-dine(3-EP). The concentration of ETS components measured rapidly decreased as increasing ventilation rate, but the removal efficiency by ventilation was different from each ETS compounds. The $CO_2$, and CO, gaseous components of ETS, were dominant components to be removed from the room by ventilation. The ventilation with 0.528 ㎥/min for 1 hr was enough to remove over 99% of those gaseous components. Nicotine and 3-EP needed the ventilation for 2 hrs to reduce over 95 % of those components. As the same ventilation rate, 99 % of TSP and solanesol concentration were removed from the room within 2 hrs, UVPM and FPM concentration decreased 90 %.

• 한국대기환경학회지
• /
• 제2권3호
• /
• pp.27-33
• /
• 1986

Atmospheric particulate matter (A. P. M.) was collected and size-fractionated by an Andersen high-volume air sampler over 15 month period from Jan. 1985 to Feb. 1986 in Seoul. The concentration of chloride, nitrate and sulfate were extracted in an ultrasonic bath and were analyzed by ion chromatography. The annual arithmetical mean of A. P. M. was 128.54 $\mug/m^3$. The concentration of anions were 2.88 $\mug/m^3$ for chloride, 3.86$\mug/m^3$ for nitrate, and 25.44$\mug/m^3$ for sulfate. The content of A. P. M. was lowest in the particle size range 1.1 $\sim 3.3\mum$ and increased as the particle size increased or decreased. And the anions exhibited a seasonal variation in the isize distribution. The contents of anions were higher in winter than summer. Ther ratio of fine particles to the total particles defined by F/T for chloride, nitrate and sulfate. The F\ulcornerT of these anion generally decrease with increasing air temperature. This tendency was prevalent in the chloride and nitrate.

• 한국물환경학회지
• /
• 제28권2호
• /
• pp.238-246
• /
• 2012

Organic matters budget in Lake Hoengseong were monthly investigated from April 2009 to November 2009. The intense rainfall occurred at between July and August and the hydrological factors were highly varied during the rainfall season. By the concentrated rainfall, the elevation, influx and efflux were sharply increased and the turbid water was also flowed into the middle water column in Lake. The inflow of turbid water increased the nutrient concentrations in water body and this appears to stimulate of phytoplankton regard as the primary productivity of influx of organic matter. Monthly average concentration of dissolved organic carbon (DOC) was generally higher than the particulate organic carbon (POC) concentration in Lake, but Temporal and spatial variation of POC concentration was higher than DOC and the maximum POC concentration was recorded in surface water in August, had the highest phytoplankton biomass. Organic carbon concentration in inflow site was rarely changed during the dry season, but the concentration was rapidly increased by the initial intense rainfall. In organic matters budget, the most of the organic matters was inflowed from the inflow site at rainfall season. Especially, the influx of allochthonous organic matters during the intense rainfall was 72.4% in the total influx organic matters.

• 한국환경보건학회지
• /
• 제24권2호
• /
• pp.134-144
• /
• 1998

A comprehensive air quality monitoring was carried out to investigate the criteria concentration of air pollutant in indoor of subway stations of Seoul City. The samples were collected twice per year (the first and the second half of the year) at each sampling point from February to September in 1997. Sampling point of subway stations was ticket office and platform. The measurement of indoor air pollutants such as sulfur dioxide(SO$_2$), nitrogen dioxide(NO$_2$), carbon monoxide(CO), carbon dioxide(CO$_2$), total suspended particulate(TSP) was performed to determine the indoor air quality. Heavy metals(Pb, Cd, Cu, Cr, As, Hg) were also measured together with those air pollutants. The annual average concentration of CO$_2$ and TSP in subway stations were relatively high while those of heavy metals were within 10% of environmental recommended standard concentration in all stations. As results of regression analysis between line and line of air factors, the concentrations of CO, CO$_2$, TSP, Pb, Cd, Cr and Cu were highly correlated, but those of $SO_2, NO_2$ and Hg were not correlated. As results of regression analysis between ticket office and platform, the concentrations of heavy metals such as Cr and Cu were highly correlated. Results of oneway analysis of variance between the first and the second half of the year air factors also indicated that CO, CO$_2$, Cd, Cu, Hg were significant($\alpha$=0.01), respectively. The average contration of total suspended particulate(TSP) in subway line No. 1 was shown high concentration(200 $\mu g/m^3\cdot$ day) level.

• 한국환경보건학회지
• /
• 제43권3호
• /
• pp.185-193
• /
• 2017

Objectives: An on-site study was conducted in order to quantify indoor exposure levels and the emission rate of particulate matter for domestic poultry buildings. Materials and methods: Three types of poultry building (caged layer house, broiler house, and layer house with manure belt) as classified by mode of manure treatment and ventilation were investigated in this study. Nine sites per each poultry building were selected and visited for measuring exposure levels and emission rate of particulate matter. Total dust and respirable dust among the particulate matter were analyzed based on the weight method. Emission rates were estimated by dividing emission amount, which was calculated through multiplying indoor concentration ($mg/m^3$), by the ventilation rate ($m^3/h$), into indoor area ($m^2$) and number of poultry reared in the poultry building. Results: Mean exposure levels for total dust and respirable dust in the poultry buildings were $3.91({\pm}1.99)mg/m^3$ and $1.99({\pm}0.89)mg/m^3$, respectively. The emission rates of particulate matter in the poultry buildings were estimated as $4.75({\pm}1.22)mg\;head^{-1}h^{-1}$ and $64.39({\pm}24.95)g\;m^{-2}h^{-1}$ for total dust and $0.58({\pm}0.23)mg\;head^{-1}h^{-1}$ and $7.52({\pm}2.51)mg\;m^{-2}h^{-1}$ for respirable dust, respectively. The distribution patterns for total dust and respirable dust were similar regardless of poultry building type. Among poultry buildings, broiler house showed the highest exposure level and emission rate of total dust and respirable dust, followed by layer house with manure belt and caged layer house. Conclusions: The finding that the broiler house showed the highest exposure level and emission rate of particulate matter can be attributed to sawdust utilized as bedding material, which can be dispersed into the air by movements of the chickens. Thus, a work environmental management solution for optimally reducing dust concentrations is necessary for broiler houses.

• 한국정보통신학회논문지
• /
• 제25권1호
• /
• pp.56-62
• /
• 2021

미세먼지 예보에 대한 높은 정확도가 요구됨에 따라 기계 학습의 알고리즘을 적용하여 예측 정확도를 높이려는 다양한 시도들이 이루어지고 있다. 그러나 미세먼지의 특성과 불균형적인 농도별 발생 비율에 대한 문제로 예측 모델의 학습 및 예측이 잘 이루어지지 않는다. 이러한 문제를 해결하기 위해 특정 농도를 기준으로 미세먼지를 저농도와 고농도로 구분하여 예측을 수행하는 등 다양한 연구가 진행되고 있다. 본 논문에서는 미세먼지 농도의 불균형 특성으로 인한 예측 성능 향상의 문제를 해결하기 위한 미세먼지 농도의 이진 분류 모델을 제안하였다. 분류 알고리즘 중 logistic regression, decision tree, SVM 및 MLP를 이용하여 PM10에 대한 이진분류 모델들을 설계하였다. 오차 행렬을 통해 성능을 비교한 결과, 4가지 모델 중 MLP 모델이 89.98%의 정확도로 가장 높은 이진 분류 성능을 보였다.

• Annals of Occupational and Environmental Medicine
• /
• 제34권
• /
• pp.21.1-21.13
• /
• 2022

Background: Because particulate matter (PM) and asthma are closely related, the prevalence of school absence among adolescents with asthma can be affected by the concentration of PM. We aimed to investigate the relationship between school absences due to asthma and the total number of days that the PM concentration exceeded the standard. Methods: We used the data from the 16th Korea Youth Risk Behavior Survey and the PM levels of 17 metropolitan cities and provinces gathered from the AirKorea. Information on the characteristics of asthmatic adolescents and the prevalence of school absence was obtained using a questionnaire, while the PM levels based on the total number of days with poor and very poor PM grades were collected from the AirKorea website. Both χ² test and logistic regression analysis were performed using the weights presented in the original dataset. Results: In the case of particulate matter of 10 microns in diameter or smaller (PM₁₀), the odds ratio (OR) after adjusting for confounders (sex, school year, body mass index, smoking history, diagnosis of allergic rhinitis, diagnosis of atopic dermatitis and city size) was 1.07 (95% confidence interval [CI]: 1.01-1.13) for absents due to asthma when the total days of poor and very poor grades of PM₁₀ (81 ㎍/m³ or higher) increased by 1 day. In the analysis of particulate matter of 2.5 microns in diameter or smaller (PM_2.5), the OR after adjusting for confounders was 1.01 (95% CI: 1.00-1.03) for absents due to asthma when the total number of days with poor and very poor PM_2.5 grades (36 ㎍/m³ or higher) increased by 1 day. Conclusions: A significant association was observed between the total number of days of poor and very poor PM₁₀ and PM_2.5 grades and school absence due to asthma; PM can cause asthma exacerbation and affect the academic life.

• 한국환경생태학회지
• /
• 제35권5호
• /
• pp.569-575
• /
• 2021

도심 교통섬과 도시숲 내부에서 미세먼지 농도와 영향 인자를 조사하여 숲의 미세먼지 효과를 분석하였다. 서울시 동대문구 홍릉시험림(도시숲)과 동대문구 청량리역 교차로에 조성된 숲(교통섬)에서 미세먼지 농도를 2018년 1월부터 11월까지 광산란법 기기를 적용하여 측정하였다. 연구 기간 동안 도시숲과 교통섬의 PM₁₀ 평균농도는 12.5 ㎍/m³, 15.7 ㎍/m³으로 나타났으며, PM_2.5의 평균농도는 6.6 ㎍/m³, 6.9 ㎍/m³으로 나타났다. 환경부 도시대기 측정망과 도시숲의 농도를 비교해본 결과, PM₁₀의 저감율은 도시숲에서 66.9±28.6%, 교통섬에서 58.6±44.1%로 나타났고, PM_2.5의 경우 71.3±23.0%, 64.9±31.3%로 각각 나타났다. 미세먼지 저감율의 차이는 도시숲의 규모와 구조의 차이와 관련이 있을 것이며, 풍속은 저감 요인으로 판단된다.

• Journal of the korean society of oceanography
• /
• 제36권3호
• /
• pp.59-71
• /
• 2001

In order to investigate the temporal variability of dissolved and particulate trace metals in the Han River, water samples were collected intermittently at two sites for 3 years (August 91 to December 94). Surface seawaters covering the range of salinity were also collected at the estuarine region to evaluate the role of estuary for the riverine fluxes of trace metals within the estuary during October 95 and 96. During the study period, dissolved metal concentrations in riverwaters varied by a factor of 5-10 for Fe, Ni, Co and Cu and 50-100 for Mn, Cd and Pb depending upon the water level; high concentration during the low water and low concentration in high water period except for Fe. The concentration of dissolved Fe increased with increasing water discharge. These concentration-discharge relationships of the studied trace metals are explained by the successive dilution of waters from two different origins, which can be presumably identified as anthropogenic discharges and watershed flushing. Although estuarine waters at early mixing region were not collected due to the difficulty of sampling, mixing behaviors of metals were inferred from the concentration-salinity relationships through the laboratory mixing experiment and field sampling, and distribution coefficients between dissolved and labile particulate phases. It is suggested that the Han River estuary plays a role of accumulating Fe, Mn, Co and Pb from riverine sources due to high turbidity caused by strong tidal current, whereas this system serves as a source of dissolved Cd due to release caused by extended residence time of riverine particles.