SECT. 2] LAKGE-SCALE INTERACTIONS 271 



component of the wind along the aircraft's flight path decreased with height, 

 and since, with the equipment used at the time, only the stress component along 

 the flight path could be measured, a negative component of the total stress was 

 recorded. As no detailed wind-profiles were available for the flight, proper 

 interpretation of the stress measurements was not possible. 



The existence of a turbulent zone between the air of the Atlantic anti- 

 cyclone and the warm sector was noted 400 km out of Bermuda during the third 

 series of observations. It was apparently non-frontal in nature ; the turbulence 

 was presumably generated by horizontal shear and changing pressure gradient 

 accelerations similar to those found in clear air in the vicinity of jet streams. 

 The heat flow and shearing stresses computed in this region appeared puzzlingly 

 large in the absence of wind data. 



In the warm sector of the cyclone all turbulence quantities dropped off to 

 minimum values for the flight, as a result of the increased stability of the air 

 moving in from the southwest. It is noteworthy that the stability overcame 

 both the tendency for increased turbulence that accompanies greater wind 

 speeds and the lateral diffusion of turbulent energy from the previously dis- 

 cussed shearing zone. The sub-normal values of turbulence lasted for only two 

 of the four observations made in the warm air. The third and fourth observa- 

 tions were made in a cool mass of air that presumably had fore-run the main 

 cold front and squeezed under the warm air. In this "cold nose" (distance 600- 

 700 km) the turbulence increased again as did the stresses, heat flow and 

 temperature deviations. The origin of the cool air is open to question since rain 

 was falling at the time of observation, which may have contributed to the low 

 temperatures directly ahead of the cold front. 



No turbulence measurements were made during frontal crossing. Although 

 the air there was rougher than normal, the turbulence was not violent ; Bunker 

 (1957) estimated the root-mean-square vertical velocity to be 150+ 50 cm/sec 

 at 300 m within this zone. The most unusual feature of the cold air-mass north 

 of the front was its abnormally high static stability, which must have been 

 maintained by strong subsidence. Beneath the level of the aircraft observations, 

 great instability must have existed since the water was 6°C warmer than the 

 air. The only turbulence observation taken in the cold air showed decreased 

 turbulence and a downward flow of heat at 100 m elevation 80 km behind the 

 front. The shearing stress measured was large and positive (3.1 dynes/cm^) as 

 would be expected since the aircraft was flying directly upwind so that the total 

 stress was recorded. 



The downward flux of heat at 100 m in an air-mass blowing over much 

 warmer water requires careful consideration concerning the balance of the 

 transfer processes at work, namely buoyant convection, mean subsidence and 

 eddy dilBFusion. Using the method of Priestley (1954), Bunker estimated the 

 maximum upflux of heat due to buoyant convection to be 0.4 meal cm~2 sec~i ; 

 this was obtained assuming the mixing rates for momentum and heat to be 

 equal and using the observed root-mean-square temperature deviations. At 

 the same time he computed a downward flow of heat by eddy diffusion of 



