winds have occurred about twenty per cent of the time during the
same months at the same levels,
The differences are due to the
high frequency of east-southeast winds at about 10,000 and 16,000
feet and the high frequency of west-southwest winds above 25,000
feet.
A slight trend for higher occurrence of winds with southerly
components is indicated as the season progresses,
Extreme care must be taken in drawing conclusions from these
data for three reasons:
~
1,
The sample is small.
The 49,000 foot data consist of less
than 100 observations for March and April,
2.
The variation of Marshall Islands weather for a given month
during successive years may be greater than the variation
during successive months,
Note the high frequency of
southerly winds at 49,000 to 50,000 feet during March as
compared to April, May, and June.
Tne weather of March,
1951, constituted most of this abnormality.
3.
The data are tabulated for each level without reference to
adjacent levels,
To evaluate the impertance of the third factor, Item 3 above,
a time-wind graph of Eniwetok winds has been analyzed for the per-
iod 1 January through 14 May 1954 (Attachment No. 2),
The winds
aloft at Eniwetok and Bikini were very similar during the entire
period except from 23 April through 5 Maye
The winds were more
southerly at Bikini than at Fniwetok during that period; and Bikini winds are shown for comparison purposes,
17h,
From this graph,