OZONE IN THE ATMOSPHERE 285 
hand, and continuous measurements of the vertical 
ozone distribution on the other hand. 
As a first possibility, one considers the meteorological 
ozone distribution as a result of horizontal advection. 
Hvery veteran observer of ozone has long been familiar 
with the fact that invasions of arctic air extending to 
high altitudes are accompanied by an increase in ozone 
amount, whereas the ozone amount is slight im the case 
of European anticyclones composed of air of southern 
origin. The increased ozone amount accompanying 1n- 
vasions of cold air has been particularly noted by 
various expeditions [5, 97]. Lejay [58] pots out that 
in the Siberian anticyclone in which the ozone amount 
is high, cold northern air masses advance so that the 
“transportation theory” removes the apparent contra- 
diction with Huropean findings. Penndorf [76] finds 
the 96-mb surface as suitable for indicating the flow 
conditions. Moser [65], in examining drawings of the 
trajectories at 5, 11, and 16 km, finds the best cor- 
relation between the trajectories at 11-km height and 
the ozone amount; he therefore assumes the main loca- 
tion of ozone fluctuations to be in the stratosphere near 
the tropopause where the maximum in wind velocity 
is located. This ozone maximum coincides with the 
lower secondary ozone layer which was first suggested, 
at least intermittently, by the investigations at Arosa 
[39]. This lower secondary layer is generally separated 
from the very constant high-altitude layer of ozone 
by a minimum in wind velocity between 15 and 22 km. 
The trajectories of Moser (Fig. 13) illustrate well the 
Con Pm Os 
TRAJECTORIES ENDING OVER: 
1=TROMSO (0.315 CM 03) 
2= AARHUS (0.258 CM 03) 
3= POTSDAM (0.339 CM 03) 
4= AROSA (0.357 CM 03) 
Fig. 13.—Air trajectories at 11 km. 
temporal constancy of ozone amount at various points 
situated in the same air current, indicating that ozone 
is a conservative property of the air and may thus serve 
as an important indicator for air-mass determinations. 
On the other hand, in Fig. 14, Moser shows that at 
Arosa the ozone amount is higher the more northerly 
the latitudes from which the air current comes. The 
source of ozone, in this case, is the ozone belt located 
by us (Fig. 7). But how is this secondary source of low- 
altitude ozone supplied? Moser and Schréer suggest 
that the great temperature gradient at the shadow 
boundary between polar night and the sun-illuminated 
0.36 POLAR 
TERRITORY 
LABRADOR - 
GREENLAND 
CM 03, AROSA 
(eo) 
ol 
ine) 
0.28 
NORTH AMERICAN BASIN= 
AZORES 
12 13 14 15 16 7 
APRIL 1942 
Fie. 14.—Ozone amount and origin of air at 11 km. 
atmosphere leads to increased shear-turbulence between 
the layer of photochemical equilibrium and the lower 
ozone layer at an altitude of 23 km. They also suggest 
that a further connection with the lower stratosphere 
is provided by high-reaching cold-core lows within which 
the barrier by the minimum in wind velocity is lacking. 
This theory needs corroboration by additional measure- 
ments. Flohn [83] sees a possible source of the winter 
singularities in the 30.5-day oscillations of the winter 
circumpolar vortex of the warm layer [43]. Craig also 
suggests continuous subsidence of ozone in the polar 
cap during the cold months. 
This leads us to the second possibility for ozone dis- 
tribution, that of transport by vertical flow. A sinking 
column of air, such as exists in low-pressure regions at 
least at the height of the tropopause, is lengthened 
(stretched) in its upper ozone-containing portion and 
this is accompanied by a horizontal convergence or 
contraction. This is equivalent to an increase in ozone 
amount. An ascending current has the opposite effect. 
Thus, the explanation of the day-to-day changes does 
not even require the assumption that sinking air comes 
directly from the layer of photochemical equilibrium 
and is there replenished with ozone. Dobson [25] 
pointed out long ago “that the air immediately in the 
rear of a cyclone often has a higher ozone value than 
in the same general air stream further to the north or 
northwest.’’? Haurwitz [50] emphasizes that horizontal 
advection cannot explain closed lines of equal ozone 
deviation such as those in Fig. 12. Moreover, the ozone 
content at the rear of a cyclone is sometimes higher 
than it is during the same season in the northern ozone 
source. Reed [81] emphasizes: 
Particularly noteworthy is the fact that closed isopleths of 
positive and negative ozone deviations bear the same relation- 
ship to surface cyclone centers as the closed isotherms which 
are invariably observed on the 200 mb chart (approximately 
