Analysis of the major factors affecting the visibility degradation in two stations.

UNLABELLED: There are four types of PM10 (particulate matter with an aerodynamic diameter <10 microm) episodes that occur frequently in central Taiwan: long-range transport with dust storms (DS), long-range transport with frontal pollution (FP), river dust (RD), and stagnant weather (SW). During the periods of the four episodes, poor visibility usually results. Multiple linear regression was applied to visibility using eight potential influential variables (temperature, relative humidity, wind speed, PM2.5, PM2.5-10, SO2, NO2, and NO) as independent variables. Of the eight variables, PM2.5 showed the greatest explainable percentage of about 48.6% and 58.1% for Taichung and Wuchi stations, respectively. This suggested that PM2.5 was the most important contributor to reduced visibility. Compared with other type of episodes, the aerosols tended to be offine size during the SWepisodes. This is the main reason that the poorest visibility occurred during the SWepisodes. Good correlation between visibility and secondary inorganic salts (NH4+, NO3, and SO4(2-)) were found at Taichung station (r = 0.71) and Wuchi station (r = 0.81), suggesting that secondary inorganic salts did contribute significantly to the degradation ofvisibility. The visibility degradation due to the effects ofNO3- was much higher than that due to SO4(2-) and NH4+ in the urban area, whereas the visibility degradation due to the effects of NO3 , SO42-, and NH4+ did not show significant diference in the rural area. IMPLICATIONS: Of the eight potential influential variables, PM2.5 showed the greatest effects on reduced visibility. Compared with other type of episodes, the aerosols tend to be fine size during the episodes of stagnant weather. This is the main reason why the poorest visibility occurred during the SW episodes. Good correlations between visibility and secondary inorganic salts (NH4+, NO3-, and SO4(2-)) suggested that secondary inorganic salts did contribute significantly to the degradation of visibility. Among the three inorganic salts, nitrates played a leading role for visibility degradation in urban areas in central Taiwan.

Aerosol characteristics of different types of episode.

Daily and hourly average data from nine air-quality monitoring stations distributed across central Taiwan, which include ten items (i.e., PM10, PM2.5, wind direction, wind speed, temperature, relative humidity, SO2, NO2, NO, and CO), were collected from 2005 to 2009. Four episode types: long-range transport with dust storms (DS), long-range transport with frontal pollution (FP), river dust (RD), and stagnant weather (SW), and one mixed type of episode were identified. Of these four episode types, the SW was the dominant type, averaging about 70%. The mean ratio of PM2.5/PM10 was the lowest during the RD episodes (0.42), while the mean ratio of PM2.5/PM10 was the highest during the SW episodes (0.64). Fine aerosol (PM2.5) and coarse aerosol (PM10-2.5) samples were collected by high-volume samplers for chemical composition analysis, from only three stations (Douliou, Lunbei, and Siansi) during the days of SW, RD, DS, and FP. The concentrations of PM2.5 and three ionic species (NH4⁺, NO3⁻, and SO4²â») all showed significant differences among the four episode types. The highest levels of NO3⁻ (12.1 µg/m(3)) and SO4²â» (20.5 µg/m(3)) were found during the SW and FP episodes, respectively. A comparison on the spatial similarity of aerosol compositions among the episodes and/or non-episodes (control) was characterized by the coefficient of divergence (CD). The results showed higher CD values in PM10-2.5 than in PM2.5, and the CD values between RD episodes and the other three episodes were higher than those between two types of episode for the other three episodes. The ratios of SOR (sulfur oxidation ratio), SO4²â»/EC (elemental carbon), NOR (nitrogen oxidation ratio), and NO3⁻/EC showed that sulfate formation was most rapid during the FP, while nitrate formation was most rapid during the SW.

Spatial variations of the aerosols in river-dust episodes in central Taiwan.

Daily and hourly average data of PM(10) from 17 air-quality monitoring stations distributed throughout four counties and along four rivers (the Ta-an River, the Tachia River, the Wu River and the Choshui River) in central Taiwan were collected from November 15 to 21, 2008. During this period, the wind speed and daily PM(10) concentrations increased drastically from the 17th to the 19th, and reached a maximum level on the 19th. Six air-quality monitoring stations (Dajia, Wurih, Siansi, Dungnan, Lunbei and Taisi) in the central Taiwan were selected for sampling fine (PM(2.5)) and coarse (PM(10-2.5)) aerosols samples during and after the river-dust episode. Most of the ratios of [crustal elements (Fe or Al)/reference elements (Cd or Mo)] for the samples obtained during episode are higher than those obtained after episode. The values of Fe(10-2.5)/Cd(10-2.5) and Fe(10-2.5)/Mo(10-2.5) were very high in Lubei, Taisi, and Dungnan stations where heavy aeolian river dust occurred. These data suggest that the ratios of Fe(10-2.5)/Cd(10-2.5) and Fe(10-2.5)/Mo(10-2.5) can be successfully used as indexes to evaluate the comparative degree of effect of aeolian river dust on the atmosphere of multiple stations.