1144 
implicit assumption that the controlling meteorological 
conditions do not change significantly during the period 
under consideration; these treatments may be termed 
then, in a certain sense, static. But the atmosphere is a 
dynamic entity whose characteristics may and often 
do change radically within short periods of time. These 
changes sometimes lead to surface concentrations which 
are not predicted by current theories. Several examples 
may be given. Dean, Swain, Hewson, and Gill [19] 
describe concentrations of sulfur dioxide found in the 
Columbia River valley in the vicinity of the smelter at 
Trail, British Columbia. During the summer the highest 
surface concentrations occur with considerable regular- 
ity at about 8 a.m., the surprising feature being the fact 
that the onset and progressive development of these 
fumigations occur practically simultaneously at sta- 
tions as far as 35 miles downvalley from the smelter. 
Bosanquet and Pearson’s discussion was available at 
the time of the investigation, but it provided no assist- 
ance in this case. The probable explanation was given 
by Hewson [38]. During the early morning hours the 
gas flows downvalley in a thin layer which does not 
reach the valley floor because turbulence is inhibited 
55. 60 55 60 55 Gol SS uIGO 
TEMPERATURE (°F) 
HEIGHT ABOV 
Fic. 3.—Various stages in the process by which contami- 
nants from an elevated source reach the ground in high con- 
centrations during a summer morning, according to Hewson 
[38] 
by the stability of the air. After sunrise, solar radiation 
heats the surface which in turn heats a layer of surface 
air, producing in it a superadiabatic lapse rate and 
marked turbulence. The thickness of this turbulent 
layer increases with time; when its upper boundary 
reaches the layer of highly concentrated gas aloft, the 
gas diffuses rapidly downward producing sudden, high, 
and nearly simultaneous fumigations along the valley 
floor. As the upper boundary of the turbulent layer 
continues to rise, upward diffusion of the gas proceeds, 
leading to exponentially decreasmg surface concentra- 
tions thereafter. The several phases involved are illus- 
trated in Fig. 3. From the theoretical standpomt a 
simplified two-stage mechanism may be visualized. 
Stage one: concentrations aloft during the early morn- 
ing are those appropriate for a continuous point source 
(the top of the stack) and a highly stable atmosphere. 
Stage two: subsequently the plume of smoke or gas 
acts as an instantaneous line or narrow area source in a 
highly turbulent atmosphere, with diffusion occurring 
initially downward and sideways but not upward; since 
the instantaneous line source moves with the wind, there 
ATMOSPHERIC POLLUTION 
is no net component of air flow normal to the line 
source. Investigations near other smelters on level ter- 
rain have shown that such simultaneous fumigations 
during the morning in summer are of frequent occur- 
rence; the effects noted above are therefore not confined 
to valleys. According to Lowry [52], experiments at 
Brookhaven show that the average surface concentra- 
tion during the fifteen-minute period with peak values 
is twenty times the maximum specified by Sutton’s 
equations. The findings at Trail, that the greatest con- 
centrations at the surface during the summer occur 
several hours after sunrise, are thus confirmed. The 
fact that high concentrations of gas or smoke may 
come to the surface in this manner miles from the stack 
must be borne in mind when peak values likely to be 
reached are of interest. 
The effect of an isothermal or inversion layer a 
short distance above the smoke plume is a second factor 
which is not allowed for by the general theory. On June 
3, 1939, a quasi-stationary frontal surface was located 
in the Columbia River valley near Trail [19, 38]. As 
shown by airplane soundings, there was a frontal iso- 
thermal layer at a level below the top of the valley 
sides but well above the top of the stacks, with a large 
lapse rate below the isothermal layer. Measured tur- 
bulence near the surface was large, but unusually and 
unexpectedly high surface concentrations of sulfur di- 
oxide were measured at stations downyvalley from the 
smelter. The low level of turbulence in the isothermal 
layer aloft prevented significant upward diffusion 
through it from occurring, and as a result the gas flowed 
downvalley as in a giant pipe. Over level terrain the 
result would not be so serious, since the gas would be 
free to diffuse laterally. However, there is little doubt 
that the presence of frontal or subsidence inversions 
and isothermal layers just above the smoke stream 
causes pollution conditions of serious concern which 
are not considered by present theories. 
It is clear that there is a great need for both theo- 
retical and experimental investigations of such special 
conditions which often lead to the most serious in- 
stances of atmospheric pollution. 
The Influence of Topography. If the terrain in the 
vicinity of a source of pollution is not essentially level, 
the difficulty of specifying what the concentration of a 
contaminant will be at a given point under various 
meteorological conditions is greatly increased. The prob- 
lems encountered when the source is in a valley have 
been fully described [19], and are twofold. When the 
winds are moderate or strong, local eddies caused by 
the configuration of the land produce a distribution of 
smoke which is well nigh impossible to predict from 
theoretical considerations. An evaluation of this factor 
requires a detailed investigation, the results of which 
can rarely be extrapolated to apply to other sites. 
Secondly, when the winds are light, local winds such as 
mountain and valley winds predominate, and even the 
prediction as to whether the wind will be up- or down- 
valley presents a major problem. The problem is further 
complicated if a number of side valleys and ravines 
lead into the main valley. The same factors, but to a 
