UPPER -WIND OBSERVATIONS OF LAST CRUISE OF CARNEGIE 



As in ordinary pilot-balloon observations, the direc- 

 tion and velocity at any minute was obtained from the 

 positions of the balloon a minute before and a minute aft- 

 er, making the usual assumption that this equalled the 

 wind velocity for the air stratum in which the balloon 

 was moving. 



Table 1 . Example of effect of steering error on 

 computed wind velocities and directions 



m/sec 



m/sec 



0.0 



Errors in the computed position of the balloon in its 

 horizontal projection arise largely from two sources: 

 (1) at the time of observation either the balloon may not 

 be centered in the eyepiece, or (2) the ship may have 

 been off her projected course. 



In the shipboard theodolite an error in centering the 

 balloon in the field of view is relatively more serious 

 than in the regular land instrument because of its larg- 

 er field of view. This larger field is required to allow 

 the balloon to be located easily. Since the field of the 

 theodolite covers 6°, the error arising from the balloon 

 being 0.1 of the radius off the center would be equivalent 

 to an azimuth error of 0.°3. The error in the horizontal 

 projection, due to an incorrect azimuth, decreases with 

 the cosine of the angle of elevation, and increases di- 

 rectly with the distance away of the balloon. With a 

 balloon at a distance of 10 km and at an angle of eleva- 

 tion of 45°, an error of 0.1 radius in centering in the 

 eyepiece leads to an error of 36 m in the horizontal pro- 

 jection of the balloon's position. 



The magnitude of the errors in the computed upper- 

 wind velocity and direction arising from the steersman 

 being unable to keep the ship exactly on her course may 

 be obtained from a short series of readings made on 

 December 30, 1928 in latitude 34.°0 south, longitude 91.° 4 

 west. For fourteen minutes an additional observer read 

 the ship's compass at the same instant as the observer 

 at the shipDoard theodolite read the balloon's elevation 

 and azimuth. The data in table 1 show that considerable 

 errors are introduced into the computed wind directions 

 and velocities by comparatively small errors in steer- 

 ing. The wind was blowing Beaufort force 4 and the 

 Carnegie according to the report was under "regular 

 sailing conditions. ' 



The largest steering error is 3°4 on the nineteenth 

 minute, which produces an error of 25° in the wind di- 



rection in the eighteenth and 0.9 m per second in the 

 wind velocity on the twentieth minute. The large errors 

 do not occur in the values for the nineteenth minute, 

 since these depend on the readings for the eighteenth 

 and twentieth minutes. It may be pointed out that errors 

 of this magnitude in wind direction occur almost entire- 

 ly with light winds. 



WIND FORCE (Beaufort Scale) 



The ship was assumed to have constant velocity, 

 which is probably correct within the limits of observa- 

 tion. 



A good helmsman may, in moderate winds, permit 

 the ship to run a quarter -point off course. During the 

 pilot-balloon flights special precautions were taken by 

 the helmsman to keep the ship on her course. During a 

 few periods of calm and adverse winds the Carnegie was 

 hove to during a flight. It was then impossible to keep 

 the ship's heading steady during the observations. Be- 

 cause of the varying azimuth, one observer stationed at 

 the ship's compass read the magnetic direction on hear- 

 ing the recorder's signal to read the theodolite. The 

 compass reading was then corrected for declination and 

 the true azimuth obtained. This procedure was found 

 tedious in practice, so that only a few flights were made 

 under these conditions. 



Rolling and pitching of the ship are generally re- 

 vealed by irregularities in the plotted positions of the 

 balloon from minute to minute. When there was a con- 

 tinuous change in the balloon's elevation for three or 

 four minutes, however, the resulting values of wind 

 shift were considered in every case to be genuine. The 

 Carnegie's periods of roll and of pitch were less than 

 ten seconds. Especially in times of light wind, if the 

 balloon was some distance from the ship, the balloon's 

 plotted positions were at times irregular, although tak- 

 en over a period of ten minutes the direction of wind 

 motion appeared definite enough. These minute-to -min- 

 ute irregularities may, of course, be owing to turbu- 

 lence, but it seemed more likely they were of observa- 

 tional origin. From this point of view, changes in azi- 

 muth readings, which led to violent wind shifts but con- 

 tinued only for a stratum of 250 m or less, were disre- 

 garded and the movement over five minutes rather than 

 one or two minutes considered. 



Considerable judgment was used in accepting the 

 data for the last minutes of a flight if they fell rapidly 

 out of line with preceding data. Frequently it would ap- 

 pear that the ship had been swinging and when it righted 

 itself the azimuth angle quickly changed its direction, 

 so that the observer failed to locate the balloon again on 

 the new course. 



