254 RADIO WAVE PROPAGATION EXPERIMENTS 
forecasting value in the sense of indicating specifically 
benefit is to be received from trapping conditions), 
these factors indicate preferable elevations of the 
transmitter. In general, the information given in Fig- 
ure 20 can be helpful in relatively long-range plan- 
ning procedures, in deciding on the most effective 
type of radar set, the optimum frequency to use, ‘and 
the most advantageous elevations at which to estab- 
lish sites, etc., for operations to be conducted during 
a particular season of the year and in a particular 
region. 
Direct Indications of Nonstandard 
Conditions in the Western Pacific 
GENERAL CONCLUSIONS 
Evidences of superrefraction have frequently been 
found at Guadalcanal and in the general vicinity of 
New Guinea and New Zealand. Data obtained on the 
radio-meteorology of the western Pacific in the region 
from New Guinea to Saipan, from soundings made 
there in 1944 and 1945,'*78 indicate the following con- 
ditions in both the equatorial and trade wind belts of 
the Pacific. 
1. Unmodified winds of long sea trajectory pro- 
duce ducts which attain sufficient height and intensity 
to trap microwaves in the 3,000- to 10,000-me range. 
2. At the relatively low wind speeds characteristic 
of the equatorial belt (4 to 10 knots), the duct height 
increases with wind speed from about 20 to about 40 
ft. X-, and possibly S-, band radars with properly 
placed antennas may be expected to show marked in- 
crease in range on surface craft and on aircraft flying 
within the duct. 
3. At the higher wind speeds typical of the trade 
wind belt (10 to 20 knots), ducts 50 to 60 ft wide 
occur regularly. Properly sited S-, as well as X-, band 
radars may be expected to show marked and persistent 
increases of range on surface targets. The persistence 
of the duct in these regions suggests its use in micro- 
wave communications. 
4. Unless modified by passage over nearby land 
masses, the atmosphere is approximately standard 
from 60 to 1,000 ft. 
5. The results are so similar to earlier measure- 
ments taken in the Caribbean on northeast trade wind 
air of long sea trajectory as to warrant the conclusion - 
that this type of duct formation is general, at least in 
tropical and subtropical regions, throughout the world. 
6. From the experimental results around Saipan 
and the Marianas it is concluded that the coverage of 
radars operated at frequencies much lower than 3,000 
me will probably not be affected by the oceanic ducts. 
7. The coverage of microwave radars (3,000 me 
and higher) which are sited above 100 ft may not be 
affected by low-level conditions. 
8. If coverage on surface craft or ultra-low-flying 
aircraft beyond the range of existing facilities is re- 
° 
74 78 82 86 13 
TEMPERATURE 
IN DEGREES F 
Ficure 21. Data from Geelvink Bay, New Guinea, 
showing surface duct as well as elevated S-shaped duct 
due to Féhn effect from 10,000-ft mountains 100 to 150 
miles to windward (SW). 
MIXING RATIO M 
G/KG 
quired and if prevailing wind speeds exceed 10 knots, 
S-band radars sited 10 to 30 ft above sea level may 
give better results than high-sited ones. 
9. X-band radars sited at 10 to 20 ft should be 
useful down to wind speeds of the order of 6 knots. 
10. In microwave communication links, the use of 
low-sited antennas may increase range beyond that 
attainable by siting at the highest available altitudes. 
(From the standpoint of water vapor attenuation and 
duct utilization, X-band frequency appears to be the 
optimum for this purpose.) 
11. From a series of ship-based kite soundings tak- 
en northeast of Saipan in the two, distinct weather 
regimes, (1) a typical fair weather period, with steady 
10- to 20-knot trades blowing and (2) a stormy period 
with variable 4- to 15-knot southerly winds and fre- 
quent rain squalls, it has been found that all sound- 
ings yield simple surface trapping curves exclusively, 
the average duct widths being respectively (1) 44 ft 
and (2) 37 ft. 
12. Although measurements were not faken under 
the conditions mentioned in paragraph 11 (2) above, 
2000 
800 
HEIGHT IN FEET 
400 
M-Moy 
Ficure 22. M deficits calculated from soundings made 
at Tateno, Honshu, 1928. 
