THE USE OF CLOUDS IN FORECASTING 
Brooks and Harwood [18] showed that heavy snow- 
storms at Blue Hill are generally preceded by Cz from 
either north or south of west, but not from straight 
west. 
Guilbert [48], in France, found that Cz come from the 
centers of depressions, that their speed indicates the 
strength but not the speed of the depression, and that 
Ci observations must be used in conjunction with the 
weather map. Peppler [76], in southern Germany, how- 
ever, found that the relation between direction of mo- 
‘tion of Cz and subsequent weather is not very close, but 
best when taken in conjunction with the pressure tend- 
ency. 
Deppermann [34], at Manila, warns that striped Cz 
usually arise from the so-called false Cz tops of Cb, 
strung out into limes by a strong upper-air stream. 
1171 
age movement of the C7, the faster the pressure systems 
move. When the temperature gradient of the tropo- 
sphere is small, however, cyclones and anticyclones may 
dominate the pressure gradients up to great heights, 
including the Cz levels. Under such conditions, the Cz 
may move from easterly directions and vary consider- 
ably in short spaces of time or distances [38]. 
Pick and Bowering [77] related 162 nephoscope ob- 
servations of Cz at Sealand and Holyhead (Britain) to 
storm tracks and tabulated the difference in degrees 
between Cz motion and the direction of advance of the 
depression center. More than half fell between 0 and 
60 degrees difference; a few exceeded 90 degrees. 
Clayton [23] and Pick and Bowering [77] have made 
statistical studies of Ci motions showing a fair, but not 
sharp, relationship to the progress of cyclones. Clayton 
TasBLe VII. Direction or DENsEST CLoups AND or CLoup Motion, aND FREQUENCY OF PRECIPITATION WITHIN 24 HR 
Cs As Ac Se 
(A) Direction in which clouds | Direction SW or N Ww WwW s 
are densest 
Cases 56 16 80 30 
Frequency of precipita- 68 81 66 80 
tion (per cent) 
(B) Direction from which | Direction NW SW SW S) 
clouds are moving 
Cases 45 16 52 ill 
Frequency of precipita- 62 69 75 91 
tion (per cent) 
Best combinations of (A) | Direction where densest SW WwW SW or W SS) 
and (B) 
Direction of motion SW or W SW or W W S, SW, or W 
Cases 30 13 24 23 
Frequency of precipita- 67 92 88 87 
tion (per cent) 
Parallel Cz give an appearance of convergence at the 
horizon. He says, “‘in quite a few cases where the writer 
thought he had convergent Cz, careful observation 
showed that there were generally two sets of parallel 
Ci of different altitudes meeting near the horizon.” 
Direction of Motion and Direction Where Densest. 
Clayton [23] noted that direction of motion and direc- 
tion of maximum density when considered jointly have 
better prognostic value than either alone (Table VII). 
The Motions of Cirrus as Prognostics of the Movements 
of Cyclones. The speed of Cz is usually a good index of 
the changeableness of the weather. Since the wind at 
the Cz level is usually dominated by the gradient of 
temperature in the troposphere, and since this gradient 
is a large element in the strength of the zonal circula- 
tion, there is, necessarily, a relation between the speed 
and direction of the Cz and the movement of cyclones 
and anticyclones. The Cz usually move from the west 
and so do the pressure systems, and the faster the aver- 
analyzed two years (1887 and 1888) of hourly observa- 
tions of high clouds at Blue Hill, 1821 in number, in 
relation to the speed of movement of cyclones, using 
mean monthly speeds of each (see Table VIII). When 
he grouped the mean monthly Cz speeds by classes of 
10 m sec and with them the storm movements, he 
found a fairly regular increase of the latter coordinated 
with that of the former, with ratios of the Ci movement 
to cyclone movement rising from 2.4 in the 15-25 m 
sec! Cz speed class to 4.5 m the 65-75 m sec class. 
It seems probable, however, that Clayton’s ratios are 
too high, because he compared average monthly Cz 
motion at Blue Hill Observatory with average monthly 
storm movement for the entire United States. Most of 
the storms, however, pass through the northeastern 
section of the country. 
Pick and Bowering, 1929 [77], using individual cases 
(162 cloud observations), found no regular relationship, 
though their tables (summarized in Table VIII) indicate 
