ATMOSPHERIC POLLUTION 
lesser degree, provide complications over any area of 
irregular terrain. Near a shore line, land and sea breezes 
occur and must be taken into account when considering 
the probable areal distribution of airborne pollution. 
The effect of topography will be considered further in 
the next section. 
The Deposition of Particulate Matter. Calculations 
of the rate of deposition on the ground of particulate 
matter from a continuous point source have been pre- 
sented by Baron, Gerhard, and Johnstone [6]. The rate 
of deposition is given by the product of the concentra- 
tion of particles adjacent to the laminar layer and their 
free settling velocity through this layer, that is, by 
xow;. The quantity xo is given by appropriate forms of 
either Bosanquet and Pearson’s or Sutton’s equations, 
and w, by Stokes’ law for small spheres with a specific 
gravity of 1.0, in air at 70F. An expression for the 
fraction ¢ of the total cloud which is deposited per unit 
time per unit distance downwind is obtained by inte- 
grating the product in the crosswind direction from 
—« to +. Thus, for example, for Bosanquet and 
Pearson’s equation, the authors obtain 
HAT of py pee 
Qo= SSE gs teN ETT é : (9) 
For Sutton’s equations a more involved analysis is used. 
The values computed with Sutton’s equation agree well 
with those for Bosanquet and Pearson’s equation for 
lapse conditions only, since the latter use an average 
value for the vertical diffusion coefficient. The essential 
results may be summarized as follows: a decrease in 
turbulence increases the maximum rate of deposition 
and shifts the point at which it occurs downwind from 
the source; an increase in stack height decreases the 
rate of deposition and shifts the point of maximum 
deposition downwind—the maximum rate of deposition 
is approximately inversely proportional to the stack 
height. A similar analysis for continuous line sources is 
available [44]. 
The Coagulation of Particulate Matter. If smoke 
particles coagulate at all rapidly in the atmosphere, 
the effect will be important in increasing the rate of 
deposition as the particles grow in size. It is known that 
coagulation by molecular movements is small [101]. 
The coagulation of smoke particles in the surface layers 
of the atmosphere has been studied both experimentally 
and theoretically by Teverovsky [88]. By means of 
ultramicroscopes the change in the size of smoke par- 
ticles in a plume was investigated: the number of 
particles and the mass concentration of the smoke 
were measured at several points along the plume and 
the average radius of the particles was calculated from 
the values obtained. It was found that with 10! to 10° 
particles per cubic centimeter the average radius in- 
creased from 0.2 » to 0.4 u while the smoke traveled a 
distance of 1 km. This coagulation is attributed to the 
action of microeddies which causes neighboring particles 
to converge. Teverovsky gives the orders of magnitude 
of various diffusion coefficients as follows: the coefficient 
for molecular movements is 10-® cm? sec™; the co- 
1145 
efficient for the microeddies to which coagulation is 
ascribed is 10-* em? sec; and the coefficient for large 
eddies which are negligible as coagulating agents is 104 
em’ sec™. The application of theories based both on 
dimensional analysis and on considerations of energy 
dissipation leads to results in good agreement with the 
observations made. These conclusions suggest that the 
possible effect of turbulent coagulation on the accuracy 
of photoelectric measurements of smoke concentrations 
should be examined. 
Wind-Tunnel Investigations. Two types of wind- 
tunnel studies have been made, those to determine the 
effect of eddies induced by the stacks themselves and 
by nearby buildings [40, 58, 73, 74], and those to 
determine the distribution of a contaminant in un- 
disturbed flow [58]. Because the effect of large-scale 
semipermanent eddies induced by buildings may be 
much more significant near the source than that of the 
prevailing field of turbulence in the atmosphere, ex- 
periments of the former type can be counted on to give 
information that is, in the main, reliable. Wind-tunnel 
studies of undisturbed flow are open to more serious 
question, however; in these it is assumed that the 
effective diffusing agency is the turbulence induced by 
the jet itself, and no allowance is made for the varia- 
tions of the natural turbulence of the ambient atmos- 
phere, which may be considerable for a given wind 
speed. Even near the stack the distribution of a con- 
taminant will be significantly affected by the degree of 
natural turbulence present. The main objection to 
studies in present-day tunnels is that the degree of 
turbulence cannot be varied so as to simulate the range 
of natural turbulence in the atmosphere which is as- 
sociated with a wide range of stability conditions. 
It is possible that various actual stability conditions 
could be simulated by using a wind tunnel in which 
the lower surface of the tunnel could be heated and the 
upper surface cooled and vice versa. Because of the 
bounding surfaces above and below and the small 
vertical extent of the air flowimg in the tunnel, it is 
probable that a lapse rate such as the dry adiabatic 
would not have the basic significance that it exhibits 
in the atmosphere. However, a lapse rate in the tunnel 
such that the air density increases with height, that 
is, a lapse rate greater than 0.034C m+, might simulate 
a superadiabatic lapse rate in the atmosphere and 
produce marked thermal turbulence in the tunnel. 
Similarly, by cooling the bottom of the tunnel and 
heating its top, strong inversions could be produced 
which would reduce turbulence to a degree character- 
istic of lesser inversions in the atmosphere. The tech- 
nical difficulties in such a study are considerable, but 
do not appear to be insuperable. Thus, inversion condi- 
tions have been obtained in the Géttingen “‘hot-cold” 
tunnel by heating the roof of the tunnel by steam and 
cooling the lower surface by running water. Detailed 
comparisons of concentrations from a given stack under 
various atmospheric conditions with those obtained 
with exact replicas of stack and surrounding terrain in 
such a wind tunnel should serve to put model studies 
on a firm basis. An excellent discussion of wind-tunnel 
