Nowadays, artificial intelligence model approaches such as machine and deep learning have been widely used to predict variations of water quality in various freshwater bodies. In particular, many researchers have tried to predict the occurrence of cyanobacterial blooms in inland water, which pose a threat to human health and aquatic ecosystems. Therefore, the objective of this study were to: 1) review studies on the application of machine learning models for predicting the occurrence of cyanobacterial blooms and its metabolites and 2) prospect for future study on the prediction of cyanobacteria by machine learning models including deep learning. In this study, a systematic literature search and review were conducted using SCOPUS, which is Elsevier’s abstract and citation database. The key results showed that deep learning models were usually used to predict cyanobacterial cells, while machine learning models focused on predicting cyanobacterial metabolites such as concentrations of microcystin, geosmin, and 2-methylisoborneol (2-MIB) in reservoirs. There was a distinct difference in the use of input variables to predict cyanobacterial cells and metabolites. The application of deep learning models through the construction of big data may be encouraged to build accurate models to predict cyanobacterial metabolites.

본 연구는 구운 횟수에 따른 죽염의 미네랄 농도변화와 항우식 효과를 확인하기 위하여 수행하였다. 미네랄 변화는 9회 죽염에서 Na, K, Mn 그리고 Fe이 높게 나타났으며, Mg, Ca 및 SO4는 1회 죽염에 서 가장 높았다. 구강 병원균 2종에 대한 최소억제농도 측정 결과 S. mutans에 대한 죽염의 MIC는 25mg/mL으로 나타났으나 S. sanguinis에 대한 죽염의 MIC는 50mg/mL로 S. mutans에 대한 항균효 과가 S. sanguinis보다 높은 것으로 나타났으며, 2종의 균주 모두에 대해 천일염보다는 죽염의 항균효 과가 더 높은 것으로 나타났다. 더욱이, 3%의 죽염에서 구강 내 pH 환경이 pH≥5.5로 오래 지속되는 것으로 나타났다. 따라서, 본 연구 결과는 죽염이 치아우식과 치주질환을 감소시키는데 유용할 것이라 생각될 수 있겠으나 보다 세밀한 추가 연구가 진행되어야 할 것으로 판단된다.

Average concentration of PM in Seoul metropolitan area satisfied the Korean air quality standard in 2010. Furthermore, concentration of PM in all boroughs across Seoul met the air environment standard in 2012. PM10 concentration was relatively higher in center of Seoul in comparison to the rest, while PM2.5 concentration showed exactly the contrary result. We analyzed the effect that PM emissions from vehicles would have on PM concentrations across Seoul. The results showed that average annual PM concentration recently decreased in Seoul although the number of vehicles registered annually continued its upward trend. By contrast, average fine dust concentrations in Seoul showed a decline which suggested that correlation between annual average PM concentrations and number of registered vehicles remained low. However, year-on-year vehicle registration rate recently showed a declining tendency in the same way as the trend of changes in average PM concentrations. Particularly, the upward trend in annual average PM concentrations in 2002 and 2007 was consistent with the increase in vehicle registration rate, suggesting that vehicle registration rate was closely associated with changes in PM concentrations.

The number concentrations and the water soluble ionic concentrations of PM2.5 have measured at Gosan site in Jeju, Korea, from March 2010 to December 2010, to clarify their characteristics. PM2.5 number concentrations vary from 22.57 to 975.65 particles/㎝3 with an average value of 240.41 particles/㎝3, which have been recorded evidently high in spring season as compared with those in other season. And the concentrations in small size ranges are greatly higher than those in large size ranges, so the number concentration in the size range 0.25∼0.45 ㎛ has more than 94% of the total number concentration of PM2.5. The major ionic components in PM2.5 are SO4 2-, NH4 + and NO3 -, which are mainly originated from anthropogenic sources, on the other hand, the concentrations of Cl-, K+, Ca2+ and Mg2+ are recorded relatively lower levels. The concentrations of the major ionic components are very high in spring season, but the concentration levels of the other components are recorded significantly high in winter season. On the other hand, in summer season, the lowest concentration levels are observed for overall components as well as the sum of them. The concentration ratios of nss-SO4 2-/SO4 2- and nss-Ca2+/Ca2+ are 98.1% and 88.9%. And the concentration ratio of SO4 2-/NO3 -(3.64) is greatly higher than the value in urban area due to no large NOx emission sources in the measurement. In addition, the correlation and the factor analysis for the number and the ionic concentrations of PM2.5 are performed to identify their sources. From the Pearson correlation analysis and the factor analysis, it can be suggested that the smaller parts(<0.5 ㎛) of PM2.5 is contributed by anthropogenic sources, but the sources of the remaining larger parts of PM2.5 are not able to be specified sources in this study.

The aerosol number concentration have measured with an aerodynamic particle sizer spectrometer(APS) at Gosan in Jeju Island, which is known as background area in Korea, from March 2010 to February 2011. The obtained results of asian dust events and non-asian dust period have been compared. The results show that the entire averaged aerosol number concentration from APS measurement during asian dust events and non-asian dust period are about 341 particles/㎝3 and 240 particles/㎝3, respectively. During asian dust events, the number concentration in small size ranges(≤0.4 ㎛) are similar to non-asian dust period, however, those in large size ranges(≥0.7 ㎛) are very higher than non-asian dust period. The contributions of the size resolved number concentration(23 channel in 0.25∼10.0 ㎛) to total number concentration in that range are dramatically decreased with increased particle size. The contributions of smaller size ranges(≤0.4 ㎛) during asian dust events are very low compared with non-asian dust period, on the other hand, those of larger size ranges(≥0.4 ㎛) are higher than non-asian dust period. total aerosol number concentration are depended on the number concentration in range of smaller than 0.58 ㎛ during non-asian dust period and asian dust events. On the other hand, PM10 mass concentration has mainly affected with the number concentration in range of smaller than 1.0 ㎛ during non-asian dust period, however, during asian dust events, the mass concentration has mainly affected with the number concentration in range of 0.65∼3.0 ㎛.

The aerosol number concentration have measured with an aerodynamic particle sizer spectrometer(APS) at Gosan site, which is known as background area in Korea, from January to September 2011. The temporal variation and the size distribution of aerosol number concentration have been investigated. The entire averaged aerosol number concentration in the size range 0.25∼32.0 ㎛ is about 252 particles/㎝3. The number concentration in small size ranges(≤ 0.5 ㎛) are very higher than those in large size ranges, such as, the number concentration in range of larger than 6.5 ㎛ are almost zero particles/㎝3. The contributions of the number concentration to PM10 and/or PM2.5 are about 34%, 20.1% and 20.4% in the size range 0.25∼0.28 ㎛, 0.28∼0.30 ㎛ and 0.30∼0.35 ㎛, respectively, however, the contributions are below 1% in range of larger than 0.58 ㎛. The monthly variations in the number concentration in smaller size range(<1.0 ㎛) are evidently different from the variations in range of larger than 1.0 ㎛, but the variations are appeared similar patterns in smaller size range(<1.0 ㎛), also the variations in range of larger than 1.0 ㎛ are similar too. The diurnal variations in the number concentration for smaller particle(<1.0 ㎛) are not much, but the variations for larger particle are very evident. Size-fractioned aerosol number concentrations are dramatically decreased with increased particle size. The monthly differences in the size-fractioned number concentrations for smaller size range(<0.7 ㎛) are not observed, however, the remarkable monthly differences are observed for larger size than 0.7 ㎛.

The aerosol number concentration have measured with an aerodynamic particle sizer spectrometer(APS) at Gosan site in Jeju, Korea, from March 2010 to March 2011. And then the atmospheric aerosol number concentration, the temporal variation and the size distribution of aerosol number concentration have been investigated. The aerosol number concentration varies significantly from 748 particles/㎝3 to zero particles/㎝3. The average number concentration in small size ranges are very higher than those in large size ranges. The number concentrations in the size range 0.25∼0.28 ㎛, 0.40∼0.45 ㎛ and 2.0∼2.5 ㎛ are about 84 particles/㎝3, 2 particles/㎝3 and 0.4 particles/㎝3, respectively. The number concentrations in range of larger than 7.5 ㎛ are below 0.001 particles/㎝3. The seasonal variations in the number concentration for smaller particle(<1.0 ㎛) are not much, but the variations for larger particle are very evident. And strong amplitudes of diurnal variations of entire averaged aerosol number concentration are not observed. Size-fractioned aerosol number concentrations are dramatically decreased with increased particle size. The size-fractioned aerosol number concentrations in size range 0.8∼4.0 ㎛ during nighttime are evidently higher than during daytime, but similar levels are appeared in other size range. The seasonal differences in the size-fractioned number concentrations for smaller size range(<0.7 ㎛) are not observed, however, the remarkable seasonal differences are observed for larger size than 0.7 ㎛.

To examine the fluctuations of aerosol number concentration with different size in the boundary layer of marine area during summer season, aerosol particles were assayed in the Ieodo Ocean Research Station, which is located 419 km southwest of Marado, the southernmost island of Korea, from 24 June to 4 July, 2008. The Laser Particle Counter (LPC) was used to measure the size of aerosol particles and NCEP/NCAR reanalysis data and sounding data were used to analyze the synoptic condition. The distribution of aerosol number concentration had a large variation from bigger particles more than 3 μm in diameter to smaller particles more than 1 μm in diameter with wind direction during precipitation. The aerosol number concentration decreased with increasing temperature. An increase (decrease) of small size of aerosol (0.3∼0.5 μm in diameter) number concentration was induced by convergence (divergence) of the wind fields. The aerosol number concentration of bigger size more than 3 μm in diameter after precipitation was removed as much as 89∼94% compared with aerosol number concentration before precipitation. It is considered that the larger aerosol particles would be more efficient for scavenging at marine boundary layer. In addition, the aerosol number concentration with divergence and convergence could be related with the occurrence and mechanism of aerosol in marine boundary layer.

To understand the development mechanism of the aerosols in the surface boundary layer, the variation in the aerosol number concentration due to the divergence and convergence of the wind fields was investigated. The aerosol number concentration was measured in the size ranges of 0.3∼10.0 ㎛ using a laser particle counter(LPC) from 0000 LST on 03 Feb. to 0600 LST on 07 Feb. 2004 at Mokpo in Korea during snowfall. The Velocity Azimuth Display(VAD) technique was used to retrieve the radar wind fields such as the horizontal wind field, divergence, and deformations including the vertical air velocity from a single Doppler radar. As a result, the distribution of the aerosol number concentration is apparently different for particles larger than 1 ㎛ during snowfall, and it has a tendency to increase at the beginning of the snowfall. The increase and decrease in the aerosol concentration due to the convergence and divergence of the wind fields corresponded to the particles with diameters greater than 1 ㎛. It is found that the fluctuations in the aerosol number concentration are well correlated with the development and dissipation of snowfall radar echoes due to the convergence and divergence of horizontal wind fields near the surface boundary layer in the inland during the snowfall.