In the literature,traffic emissions and/or regional transport of NH3 associated with agriculture activities have been proposed to be two major contributors to NH3 in urban atmosphere.However,evidences were limited in these studies.In this study,a Dionex Gas Particle Ion Chromatograph was used to semicontinuously measure NH3 gas and NH4 + in PM2.5 in December 2008 at a roadside site in downtown Toronto,Canada.In December,regional emissions of NH3 associated with agriculture activities were much low because of frequently snowing and low ambient temperature (T).In addition,the semi-continuous measurement demonstrated that the contribution of on-road vehicle emissions to atmospheric NH3 was less than 1 ppb,consistent with the observations nearby a highway presented in the companion paper.However,the daily maximum mixing ratio of NH3 in ~50% of days was observed at 4:30 pm -7:00 pm and ranged from 3.4 to 9.1 ppb.The maximum value in these days was generally associated with the low wind speed and overwhelmed from emissions inside the street canopy.Upon further examinations,emissions from the grassland nearby the site were likely an important source of atmospheric NH3 when T was over 2℃.There was ~30% of days in which the daily maximum value of NH3 occurred at 4:30 pm -7:00 pm,but T was below 2℃.These spikes of NH3 in the 30% days were unable to be explained by known emission sources,implying that major sources of NH3 in the urban atmosphere in winter were still unknown.