dense set of longitudes will have been sampled, ostensibly allowing resolution 
of comparably fine structures which are slowly evolving over the sample 
period. Another intriguing aspect concerns the latitudes where ascending and 
descending nodes coincide. The number of longitudes sampled at these latitudes 
is half that of other latitude circles. In this case, it would appear that 
the zonal resolution is also halved. It will be shown that, because of the 
nonconcurrent nature of asynoptic measurements, both of these arguments are 
fallacious. 
2. Sampling and Orbital Geometry 
sampling patterns for polar orbiting meteorological satellites 
encircle the globe every 24 hr. The trajectory of observations forms a 
sample net, shown in Fig. la for nadir viewing instrument and in Fig. lb for 
a 30? (from the orbital plane) limb scanning instrument. For both cases, the 
longitudinal spacing between ascending and descending nodes, AA.,, is uneven 
(£ 180?), varying with latitude and approaching zero near the eÂfremeties of 
the orbit. We proceed now with observations on a single latitude circle. 
  
(a) @ Ascending nodes 
O Descending nodes 
   
   
   
    
  
     
   
   
   
K £^ 
SEE 
UV 
WR qu 
Y A 
ARR 
> / X y) NY 
      
AP 
det 
A d. 
M A CA 
ASS 
   
   
     
   
  
    
  
  
     
  
  
Les 
  
    
  
  
    
    
    
     
     
   
   
  
   
s 
hil 
t XX T | 
     
  
  
  
  
  
  
Fig. L Observational "net" for 1 day's cycle. a) nadir sonde, b) 309-limb scanner (Vo = 13.8 orbits/day). 
Coordinates of the jth ascending node are given by 
Aaj = jest, (1.1) 
taj m5jto? (172) 
where 
di, 
Co = dt - - 2m ’ (2.1) 
^ is longitude, t is time (days) and c, and c. are the orbital period and 
speed that the orbital plane precesses"about the latitude circle. The orbital 
period is related to the orbital frequency, Yo (orbits/day) as 
0 
at E (2.2) 
156 
NN EN EE ON 
P, 
MEME en ON” 
Simi 
wher 
asce 
zero 
conv 
desc 
the 
(Fig 
desc 
locu 
T/2]. 
doub]