SATELLITE-TRACKING PROGRAM — HAYES 319 



There was, however, one problem. No one knew what values to 

 insert into the equations for the flatness of the earth; consequently, 

 one could not predict the rate of rotation of the orbital plane. 



Using the numerical integration procedures developed by Cunning- 

 ham, Dr. Lautman worked out a second computer program. This 

 required taking a set of initial conditions, carrying the numerical 

 integration forward step by step to the times at which observations 

 were available, computing the observations that would be seen if the 

 orbit were correct, comparing those with the actual observations, 

 and computing the differences, that is, the residuals. This is done 

 for one estimated orbit and a dozen observations, from which are 

 derived a dozen sets of errors or discrepancies. Then, one by one 

 the elements of the assumed orbit are corrected, and the corrections 

 are applied to the original orbit, new predictions are computed, the 

 predictions are then compared with new observations, and a second 

 set of residuals is determined. A statistical study of this ensemble 

 of residuals leads to an estimate of the "best orbit." 



The program developed by Lautman could not be used for making 

 practical day-by-day predictions. First, it required a very good 

 initial guess at the orbit. If the guess were poor, the next guess might 

 be even worse, and impossible predictions would result. Second, 

 the calculations for each improvement of the orbit were very time- 

 consuming on the IBM-704 electronic computer. As much as an hour 

 might be required for each stage of the improvement. 



Late in the summer of 1957 the Observatory staff decided that they 

 would probably require a third computer program which would 

 permit the computation of an orbit from three observations. This 

 would be especially needed since the differential correction programs 

 demanded an initial orbit that was fairly accurate. Jack Slowey 

 modified a method developed primarily by Robert Briggs to include 

 the secular variations in the orbital elements due to the earth's bulge. 

 Working closely with John Rossoni of IBM, Slowey and Lautman 

 were debugging the program early in October. 



COMMUNICATIONS 



In the summer of 1957 Norris D. Pease, a communications con- 

 sultant, laid out the plans for the communications center at 79 Garden 

 Street, Cambridge. His objective was to provide means of communi- 

 cation between Cambridge and the satellite-tracking stations and 

 the Moonwatch teams and to coordinate these so that there would 

 be rapid transmittal of information. Through this network the com- 

 munications center could send preliminary orbital data to the Moon- 

 watch teams, the teams could make observations and send determina- 

 tions of time and position to Cambridge; Cambridge then could 



