The paper presents an analysis of a set of data that includes both Hg profi
les and modeled fluxes over water surface (using the two-layer model) and d
evelops some insights regarding mercury water-air exchanges in Lake Ontario
and the Upper St. Lawrence River (USLR). This paper presents the first in
situ estimation of the water-air Hg gas exchange in Lake Ontario. It shows
that water-air exchange fluxes over Lake Ontario and the Upper St. Lawrence
River contributed to atmospheric mercury built-up over water bodies. TGM c
oncentrations over Lake Ontario were significantly higher on the South shor
e than on the North shore, whereas the degree of saturation was higher on t
he North shore. Moreover TGM concentrations over water bodies were signific
antly higher than the reference station located on the ground surface (St.
Anicet) indicated. TGM gradient measurements suggested that the South shore
of Lake Ontario was mostly in evasion mode, whereas on the North shore and
the Upper St. Lawrence River, bath evasion and deposition were observed. H
owever, Hg gas evasion should be far larger than gas deposition. Hg fluxes
were calculated through the two-layer model. Model calculation indicated Hg
supersaturation (476-2163%) and Hg fluxes between similar to 0 and 9.28 ng
/ m(2)/h (median 2.88 ng/m(2)/h). Hg gradients were weakly related to solar
radiation, whereas the calculated Hg fluxes were not. Modeled Hg fluxes ar
e related to Henry's law and wind speed. It is suggested that the two-layer
model gives the order of magnitude of the Hg flux but cannot express adequ
ately the fine structure of Hg water-air fluxes. Further research is needed
to elucidate the fine structure of the Hg water-air gas exchange flux. Thi
s study points out the complexity of Hg water-air gas exchange flux process
es over large lakes or rivers acid that Hg gas exchange is dynamic, changin
g in space and time.