AIR TEMPERATURE 25 
Table 30. Corrected values of Fourier coefficients, amplitudes, and phase angles of the 
24-hourly and 12-hourly oscillations of air temperature, Carnegie, 1928-292 




Desig- 
nation 


Latitude range and number of days used in determining corrections 


Coefficients, °C 
aj -0.390 -0.202 -0.247 -0.371 -0.293 -0.429 
aQ + 0.056 +0.057 -0.014 +0.106 -0.036 +0.110 
bj -0.159 -0.096 -0.366 -0.234 -0.162 -0.117 
b2 -0.097 +0.051 + 0.002 +0.004 -0.005 -0.034 
Amplitudes, °C 
cl 0.421 0.224 0.442 0.440 0.335 0.445 
c2 0.112 0.082 0.014 0.106 0.036 0.115 
Phase angles, ° 
o1 247.8 244.6 214.0 237.8 241.1 254.7 
$2 150.0 48.2 278.1 87.8 262.1 107.2 

Coefficients, °C 
al ‘ -0.274 -0.429 -0.442 -0.453 -0.460 -0.3772 
ag +0.072 +0.094 +0.106 +0.024 +0.035 +0.0925 
bi -0.111 -0.128 -0.119 -0.197 -0.362 -0.119 
b2 -0.038 -0.046 -0.147 -0.042 +0.027 -0.039 
Amplitudes, °C 
C1 0.296 0.448 0.458 0.494 0.585 0.395 
C2 0.081 0.105 0.181 0.048 0.044 0.100 
Phase angles, ° 
1 247.9 253.4 254.9 246.5 231.8 252.5 
$2 117.8 116.1 144.2 150.3 52.4 113.0 
De ee 
2 Corrections obtained from difference between dry-bulb readings in screen on deck and at 
the crosstrees. 
These are means for latitude zones 15° N-5° N, 5° N-5° S, and 5° S-15° S, and from these 
amplitudes, c, and phase angles, ¢, were determined. 
air temperature for excesses during daylight hours in 
tropical regions of the Pacific Ocean (between latitudes 
+15°) decreases the amplitude c2 by almost one-half and 
increases the phase angle, $9, by 27°. The maximum 
amplitude thus occurs almost one hour earlier. 
It now appears that previous measurements of air 
temperature on board vessels at sea have been too high, 
at least by several tenths of a degree, and that ampli- 
tudes obtained from the uncorrected temperatures over 
the ocean are in error. If this is the case, then certain 
theoretical considerations regarding the dynamics of 
these oscillations, such as Chapman’s [27], will need 
modification. 
From data obtained during the Meteor Expedition 
(1925-1927), Kuhlbrodt and Reger [28] found the diurnal 
variation of air temperature to be of the order of 0°3, 
somewhat smaller than is indicated by the Carnegie am- 
plitudes for cj. 
The diurnal variation of temperature is large com- 
pared with the semidiurnal variation, in direct contrasi 
with the case of pressure. In the latter case, the ampli- 
tude of the 12-hour wave is greater than that of the 24- 
hour. Within latitudes +20°, the Carnegie mean ampli- 
tudes of the 12-hour waves of pressure average roughly 
three times the amplitude of the 24-hcur wave. The dif- 
ference in phase between the 12-hour pressure wave and 
the 12-hour temperature oscillation averages approxi- 
mately 72°, which is equivalent to stating that the time 
of maximum pressure occurs 2.4 hours earlier than the 
time of maximum air temperature. For uncorrecied 
data (table 27), the phase differences are more irregu- 
lar, but between latitudes +20° the time of maximum 
pressure averages only 1.5 hours earlier than the time 
of maximum air temperature. 
Bartels [9, pp. 17-19] has considered the relations 
between the 8-hour terms of pressure and temperature 
for certain European stations. Chapman [27], Pramanik 
[19], and Topping [29] have collected and analyzed data 
for several land stations during January and July. They 
found the phase to be fairly regular in summer and the 
amplitude generally greater in winter. Bartels found 
the amplitude of this temperature wave for Potsdam to 
be least at the equinoxes. The Carnegie coefficients of 
the 8-hour temperature oscillations are so irregular that 
they cannot be said to verify any of the above conclusions. 
The amplitudes and phase angles of this wave as given 
by Chapman, Pramanik, and Topping for Mauritius and 
Ascension for January are: Mauritius (20° 06’ south), 
c©3= 0°35 C, ¢3=35°; Ascension (7° 55’ south), c3 = 
0°38 C, ¢3 = 45°. According to the Carnegie observa- 
tions, the amplitude of this oscillation is considerably 
smaller than the values given above. The Carnegie ob- 
servations for mean latitudes 15° to 25° south (31 days, 
November and January - March) and for mean latitudes 
