am unable to get better than 4 jm pointing accuracy in the corners where pass points fall. The deformation 
pattern remains approximately the same for one camera/magazine combination, particularly the larger 
residual vectors. (See his Invited Paper in Commission I). 
J.B. Case. I support the iterative procedure with analysis of residuals, but in the first iteration the ground control 
should receive a high weight, or systematic errors will go into the control co-ordinates. 
H. Ebner. The effect of systematic errors is very dependent on the bridging distances between controls. If 
systematic errors are introduced into the stochastic model, a very big covariance matrix results; they can also 
be introduced as weighted observations with expectations. In this way the programme is simpler and more 
general. 
K. Kraus. Kubik suggests use of correlation theory. I have done this for two years, and have published a full 
description. It is used to correct triangulation after adjustment. Better homogeneity of the maps results, 
especially with sparse control. There is a presented paper available (Paper III.18 with E.M. Mikhail). 
M. Cunietti. We need to look at the physical aspects of errors. I believe that it is possible to determine the trend 
and nature of systematic errors by studying the strips themselves, without needing too much ground 
control. 
H.H. Schmid. The sources of systematic errors are of concern to both Commissions I and II, but if they cannot 
be eliminated, Commission III must seek ways to correct for them. 
Friday August 4 1972, 10:45 
Invited Paper: **Measuring the Moon with Apollo 15"' by L.A. Schimerman. 
Panel discussion 
F.J. Doyle. A control network, established from earth-based astronomical photographs, has an accuracy that 
varies from 1/2 km to 3 km from the centre to the periphery of the lunar disc. There are differences of up to 
100 m between various libration models which indicates a significant uncertainty. 
The relative timing of photographs in one pass is good to about 1 m sec, the relative times between 
adjacent passes to 5 m sec, and the overall accuracy on UT is only about 20 m sec. The lack of Doppler tracking 
data, when the spacecraft is behind the moon, also reduces the accuracy of the ephemeris. It is unlikely that 
this photography will significantly improve the knowledge of lunar long (several years) and short (1 month 
or so) period librations, or gravity field, owing to the very short coverage and short isolated periods of 
photography. 20 Apollo missions were planned, the last six to have metric cameras; the last three have been 
cancelled. Apollo 15, at a 30 ^ inclination, only covered about 1046 of the surface. Apollo 16 had an inclination 
of 10%, and much of its photography duplicates that already obtained. Apollo 17 is planned for Dec. 1972; 
80% of its photography will duplicate that of Apollo 15 and 16, and the three together will achieve 16% 
coverage. There are hopes for a photographic mission in a polar orbit, but if this takes place, it will not be 
before 1980. 
H.H. Schmid. Ranging to the emplaced corner cube reflectors has a precision of about 30 cm, and the data can 
be fitted to the lunar motion to about X10 m. This is a continuing programme which should result in 
improving knowledge of librations. The photographic problems are basically those of timing and the 
inadequacy of the coverage, because geodesists have only a minor influence on mission planning. 
D.W. Proctor. Although today's speakers regret the low coverage, and cannot greatly influence planning, they 
are personally largely responsible for the great improvement in the metric properties of lunar photography 
since the days when cameras were used in which neither the lenses nor the reseau had been calibrated. Another 
significant point is the example to terrestrial photogrammetrists of the use that can be made of auxiliary data 
when the need arises. 
E.H. Thompson. One source of this data is not normally available to terrestrial photogrammetrists; the 
simultaneous use of a stellar camera. 
FJ. Doyle. There are plans for earth-orbitting vehicles with metric cameras linked to stellar cameras, and 
supported by many tracking stations. Such a system should be able to establish control suitable for 1:25,000 
mapping in those areas where no such control already exists; the techniques developed for the Apollo missions 
are capable of doing this.