LAUSR

Abstract The NH3(g) concentration measured at multi-site network in Kobe and Kitakyushu, Japan by using passive method continuously for one year in the late 2000s (September 2008–August 2009) and March 2018–February 2019, respectively, was statistically analyzed from the viewpoint of its controlling factors for temporal variation and spatial distribution. Nine survey sites were established in a region of Kobe, where highly urbanized, agricultural and mountainous areas were located within a 20-km x 20-km region; and one site in Kitakyushu, considered as urbanized and industrialized area, was taken into account. In terms of temporal variation, in both areas of Kobe and Kitakyushu, the NH3(g) concentration indicated lower level in the summer than those in other seasons and experienced higher concentration in colder period, which was newly found compared with the conventional wisdom and results in former studies. Regarding spatial distribution, the NH3(g) concentration showed large variations depending on the site, and the annual mean NH3(g) concentration ranged from 0 ppb (not detected: N.D.) to around 5 ppb. The NH3(g) concentration at the mountainous site with high elevation was the lowest, while it was the highest at the site where the contribution of agriculture was noticeably large in the NH3 inventory, followed by commercial and residential areas. The meteorological factors significantly contributed to the specific seasonal variation; moreover, multiple factors such as located situation, acid-base balance in the atmospheric reaction and vehicular emission were mutually related and presumably responsible for the seasonality of higher in the colder season and lower in the warmer season. When the NH3(g) concentration in the urbanized area is discussed, as in our present study area, the proportion of humans and pets (Human&Pet) to the total in the NH3 inventory could be a good parameter to account for the NH3(g) concentration.