218 MALKUS [chap. 4 



are strong, zonally uniform and relatively uninterrupted by the invasion of 

 disturbances. These "high index" periods alternate with those of "low index" 

 when the subtropical ridges are weaker, elongated meridionally and often 

 broken up into two or more feeble centers over an ocean basin. Under these 

 conditions, disturbances are common and the trade flow is weak and disrupted. 



Thus, predominant speeds above 6-7 m/sec will normally be associated with 

 high index, lower speeds with low index. Further, at the start of high index, 

 when the surface ridge builds in the subtropics, we observe strong acceleration 

 of the trades with flow toward lower pressure, i.e., equatorward. At times this 

 acceleration is spectacular resulting in "surges of the trades" with speeds of 

 15-20 m/sec over wide portions of a tropical ocean. It is likely that the equator- 

 ward flow is greatest during the period of acceleration. Also the rapid transport 

 of relatively cool dry air over warm water should cause evaporation to reach 

 a maximum at these times. High index, however, is also the period of minimum 

 exchange of air-masses across the subtropics. After an initial surge of perhaps two 

 to four days' duration, we can expect that the temperature difference between 

 sea and air will again decrease, initiating the next stage of the index cycle. 



These oscillations were first noted by Riehl (1954a), who has emphasized 

 their importance to the mid-latitude energy supply. Subsequently, Kraus 

 (1959) attempted a theoretical analysis of their time-scale in terms of the delay 

 between evaporation in the trades and precipitation in the equatorial trough. 

 A weak link confronting further progress lies in our vast lack of knowledge 

 concerning the stability of the easterly trades to the formation and deepening 

 of synoptic disturbances. If it were established that the instability to dis- 

 turbances became greater as the trade current decreased, a physically reason- 

 able chain of events is suggested ; unfortunately, so many interacting circulation 

 branches must be considered that it is naive to hope for a simple or rapid 

 resolution. 



In itself, the observation that the trade strength may vary by a factor of 

 two or more about its seasonal mean is significant as far as export of latent 

 heat is concerned, particularly since the time scale appears to be a crucial one 

 in meteorology. These variations become even more significant when we re- 

 examine Fig. 6 and find that the mean speed, about which the fluctuations 

 occur, is in the region of strongest slope of the Cd curve, so that amplified 

 variations in exchange of energy and momentum may result simultaneously 

 over wide regions. It is probably not coincidental that 6-7 m/sec is also the 

 critical air velocity for the formation of white caps. The role of wave shape and 

 sea spray in transfer processes has been discussed but not resolved (Munk, 

 1947; Neumann, 1951; Rofl, 1952). 



The consequences to the ocean, however, of the most conservatively esti- 

 mated evaporation fluctuations are not negligible. Reducing it by one-half over 

 a month retains about 3000 cal per cm^ of sea surface that would otherwise be 

 removed by the atmosphere. If the upper 100 m share this equally, the mean 

 temperature of this layer would become 0.3°C above normal : a small-sounding 

 increment until we recall that it is comparable to the average difference between 



