objectives. However, the edge location of fiducial marks is in the 
area of greatest film distortion and the described dynamic condition 
of the reseau and fiducial systems requires extrapolation of defined 
film distortion in relating image measures and reseau marks to the 
fiducials. In practice each image and its four surrounding reseau 
marks and each fíducial and its adjacent reseaus are measured and 
adjusted into the calibrated reseau system. A final adjustment of 
measured to calibrated fiducial values places all points in the 
e 
fiducial system. Standard error of these last adjustments is 
typícally 5-6 microns, indicative of effect of defined residua! film 
distortion. The net result of this condition is an uncertainty of 
approximately 3 microns in defining principal point location, which 
in the triangulation process, will probably become inseparable from 
exposure station position error. 
Relative analytical triangulation assemblies of the three test 
strips demonstrated plate residuals at a 5-10 micrometer level, 
corroborating the geometric integrity of the terrain photographs 
upon successful application of film and lens distortion corrections. 
The principal source of variance in dimension of plate residuals 
was photographic resolution as affected by sun angle. Increase in 
residuals within the 5-10 micrometer range was closely correlated 
to increases in the sun's elevation in progressive orbital revolu- 
tions. Sun elevation angles of 13?-20?, 17?-24? and 44?-52? were 
recorded respectively for revolution 22, 27 and 60 test frames. 
Results were only slightly improved (1-2 micrometers) where solution 
was computed using successive rather than alternate photographs in 
& strip. 
The Apollo 15 analytical triangulation is dependent upon 
reduction of stellar photography to derive terrain photo orienta- > 
tion through the calibrated 96 degree interlock angle between 
stellar and terrain cameras. Essentially, the process entails 
mensuration of stellar, surrounding reseau and fiducial images 
to provide corrected coordinate values related to the stellar 
photo principal point. Based on identification of three or more 
stars and time of exposure, stellar coordinates are first computed 
in the celestial system and then terrain camera orientation is 
derived in the selenocentric inertial and local lunar vertical 
- systems through a series of mathematical rotations. Included as 
input are the locking angle and Koziel's lunar libration model. 
Good results are being achieved in reduction of Apollo 15's 
stellar photography with 20-40 stellar images/photograph normally 
being recovered, allowing derivation of orientation with a 
precision of 20 seconds of arc. 
TI 
to lun. 
an unu: 
laser 
be ima; 
of poii 
positi: 
Incorpe 
require 
define: 
locatic 
of the 
statior 
derivir 
precis: 
lation: 
contril 
diagnos 
altimei 
photogi 
are ave 
frames 
level 
determi 
Comparj 
differ: 
TI 
is pro 
vector 
affords 
definec 
selves 
knowlec 
observ: 
of expc 
revo lut 
an errc 
statior 
of timi 
implyir 
photogr 
error ij 
positic 
gravity 
the moc 
to defi 
raphy w