SATELLITE-TRACKING PROGRAM — ^HAYES 339 



which time tlie satellite may perform as many as 150 revolutions 

 around the earth. During this period, any errors made by not com- 

 puting enough significant figures tend to accumulate. 



Cunningham's effort was aimed at constructing a program that 

 could be used as a standard reference for computing accurate, defini- 

 tive orbits after all the observations were in, and for checking more 

 approximate theories. His work was not intended to provide a prac- 

 tical approach to computing orbits on a day-to-day basis, for his 

 program required at least one minute to compute a single orbit of 

 perhaps an hour and a half. 



By mid- 1958, the program was being debugged and checked out. 

 At the same time, it was being modified so that elements of it could 

 be included in the differential correction program of Dr. Lautman. 



The latter program had been completed by late 1958, thus providing 

 an extremely accurate method of correcting orbits of satellites, with 

 or without drag. Again, the large amount of computer time necessary 

 for its operation precluded its use for day-by-day corrections and 

 ephemerides. The Observatory expected, however, that its accuracy 

 and general applicability would result in its use as a powerful tool for 

 analysis, especially when geodetic satellites were available. 



Both of these programs required that the magnitude of satellite 

 drag, the size and shape of the satellite, and other physical parameters 

 be known and included in the calculations. In contrast, a differential 

 orbit improvement program developed by Dr. George Veis included 

 virtually everything as unknown and approached the problem purely 

 as one of defining the orbit without having recourse to theory. 

 The theory came afterward once the motion of the satellite had been 

 detennined. 



The Russians had developed such a program, which seemed the most 

 practical way to compute orbits for generating predictions. Mean- 

 while, Veis had included in his doctoral dissertation at Ohio State a 

 chapter on satellite-orbit computing that contained all the features of 

 the Russian program. His method was entirely satisfactory from 

 every point of view. He had worked it out independently and had not 

 the slightest notion of what was being done elsewhere. When Dr. 

 Whipple learned of the features of Dr. Veis' program, he asked that it 

 be set up as quickly as possible for use at the Observatory. 



Dr. Veis' program had originally been developed for geodetic pur- 

 poses, that is, he planned to use it to determine precisely the positions 

 of stations from which observations of satellites were made. The 

 problem now was to invert that program in such a way that, the posi- 

 tions of the Baker-Nunn stations being relatively well known, the time 

 and position of satellites could be determined from observations made 

 from those stations. 



