220 MALKUS [chap. 4 



have ever been captured for quantitative study (Malkus and Ronne, 1954). 

 Concentrated in rows in the convergent area of a polar trough disturbance, 

 these shot upward into the intense westerhes of the high troposphere, their tops 

 spreading in 100-km streamers in the jet, centered at about 45,000 ft. Measure- 

 ments from time-lapse films showed towers several kilometers across rising 

 from a 7-10 km wide cloud body over open ocean at ascent rates exceeding 

 12 m/sec. Simultaneously, the nearby trade cumuli were sparse, feeble and 

 stunted. A computation similar to that of equations (46-50) showed that, 

 averaged over the convergent area, the net upward vapor transport by the 

 cumulonimbi was not less than 400 cal/cm^ per day, or nearly twice the normal 

 evaporation rate. 



Thus the ordinary trade cumuli seem to require a good mixed layer, wind 

 stirring and the normal operation of exchange, but grow well in the face of the 

 weak subsidence ( ~ 10"^ per sec ; see Fig. 46b) characteristic of the undisturbed 

 trade, while giant cumulonimbi are dependent mainly on intermittent flow 

 convergence. Over the open sea, the latter is associated with the passage of 

 synoptic-scale disturbances. The frequency of these depends on the stability 

 of the air flow, which varies with location and season. In regions where they are 

 common, it is unlikely that the dynamics or energetics of air-sea interaction 

 can be modelled or understood mechanistically without taking this scale of 

 motion carefully into account. 



b. Effects of moderate disturbances and the diurnal cycle 



{i ) The nature of synoptic-scale perturbations in low latitudes 



The only type of tropical disturbance known to most mid-latitude dwellers 

 is the notorious hurricane, which they would prefer remained at home in the 

 tropics. In addition to its destructive record and distant excursions, the 

 hurricane is the real black sheep among tropical disturbances on several other 

 counts, the most important being its rarity. An average tropical station ex- 

 periences ten or more synoptic-scale perturbations per month ; less than 10% 

 of these, even in season, attain the dubious fame of a girl's name. In addition, 

 the hurricane extends through the entire troposphere and, as any schoolchild 

 knows, intense winds (by definition > 65 knots) rage about its center. 



In the latter sense, most tropical disturbances are not properly termed 

 "storms" since they commonly exhibit weaker-than-average winds until 

 moderate development is exceeded, nor until then do they extend throughout 

 the troposphere, in general being confined principally to its lower or upper half. 

 As has been brought out previously, the significant feature of the tropical 

 atmosphere is its two-layered structure ; ordinarily the different disturbances 

 of each layer move within it, at speeds comparable to its mean wind, and to a 

 degree, independently of what goes on in the other. In fact, although little is 

 known of the formal criteria, the requirement for an intense storm is its ability 

 to lock together the layers, creating a coherent circulation which occupies 



