224 RADIO WAVE PROPAGATION EXPERIMENTS 
earth. Within these ducts these waves may be trapped, 
and abnormally long ranges may occur. 
As the trade winds blow toward the equator over 
warmer ocean areas the heating from below causes 
the duct to rise and to become weaker until finally 
near the equator the duct disappears and the two 
air masses become thoroughly mixed. 
HEIGHT OF THE Duct BAsE 
The base of the inversion (or duct) increases in 
elevation equatorward. According to the Meteor 
soundings, taken during March and April, the aver- 
age elevation of the base of the inversions rose from 
700 m at latitudes 15°N to 20°N, to 1,020 m at 10°N 
to 15°N, and to more than 2,000 m at latitudes 5°N 
to 10°N. Between the equator and 5°N, no ducts 
existed to an elevation of 2,500 m. 
The elevation of the base of the inversion also in- 
creases westward into the Atlantic from the African 
Coast. At latitudes 15°N to 20°N, its elevation in- 
creases from less than 300 m off the African Coast 
to 1,500 m in mid-Atlantic. 
METERS 
50 40 
60 To 60 
Ez. 
Tact 
if 
NY ae 
a i 
4 
Galwanisiy 
eae 
Ficure 13. Height of the temperature inversion base 
over the Atlantic. (After von Ficker.) 
Figure 137+ depicts the height of the temperature 
inversion base in the Atlantic, based largely on the 
data from the Meteor expedition. South of the equa- 
tor, soundings were made during the winter season 
(June to August), while north of the equator the 
soundings were made chiefly in the spring (March 
to May). The height of the base of the inversion has 
a seasonal variation, being greater in winter than in 
summer. 
Figure 14 represents typical M curves computed 
from the soundings of the Meteor expedition. Curves 
A (sounding 182 of the Meteor expedition taken just 
off the African Coast) show a ground-based duct of 
elevation 140 m on the ascent curve and 90 m on the 
descent curve. Oceanward, the duct becomes ground- 
based, S-shaped, as is shown by curves B (sounding 
183) 
300, 
HEIGHT IN METERS 
CURVES A AND B 
CURVE C 
HEIGHT IN METERS 
0. 
310 «©3300 «= 3350 310 
Ficure 14. M curves, Meteor expedition, March 1927. 
The descent curve! shows a decrease of 59 M units 
in 20 m, corresponding to a ray curvature about 20 
times greater than that of the earth’s surface. West- 
ward into the Atlantic the inversion base rises to about 
1,000 m. (Curve C, sounding 184.) 
FREQUENCY oF Duct OccURRENCE 
Within the trades proper a duct is practically cer- 
tain to exist. In Table 14 the percentage of duct 
occurrence, by latitude according to Meteor sound- 
ings, is tabulated. The extreme curve (ascent or de- 
scent) was utilized in determining the existence of a 
duct. 
TaBLE 14. Frequency of duct occurrence by latitude 
over the Atlantic Ocean based on the Meteor soundings. 
Latitude, degrees Frequency (% of all cases) No. of cases 
20-15N 100 19 
15-10N 7 19 
10-5N 40 17 
5-0N 0 18 
0-58 0 17 
5-12S 56 16 
12-18S 53 15 
18-248 73 26 
24-358 10 20 
35-508 19 21 
50-638 0 20 
It is evident from Table 14 that in the vicinity of 
the doldrums (5°N to 5°S) the existence of ducts is 
rare. A maximum frequency occurs between latitudes 
15° and 25°. Note that all soundings made between 
latitudes 15° to 20°N indicated the presence of a 
duct. 
THICKNEss oF Ducts 
The average thickness of the duct in the trade 
wind inversion, according to Meteor data, is about 
130 m. According to the theory of the dissipation 
of the ducts near the equator due to heating from 
below; the thickness should decrease toward the equa- 
1The ascent and descent curves disagree largely because of 
the lag of the humidity element in the sounding rig. The curve 
showing sharper inversion is therefore probably the more 
accurate. 
