= 
i 
to be a practical limit. Table No. 2 shows that for an aerial observer both 
the absolute amount of refraction and its differential change have their maximum 
values for a flying height of about 18 km. For a depression angle of 35° the 
corresponding values are 27" and 1", respectively. Based on practical experi- 
ence with metric aerial photography, one should be satisfied with a compu- 
| tational accuracy of a3 for the refraction correction. Consequently, the 
corresponding depression angles for the individual rays can be computed by 
  
simple formulas depending on & local Cartesian coordinate system, provided that 
i its origin is within 41? = — +100 km from the camera position. This approach 
| will do justice to strips of 200 km length or biocks with sides of about 150 km. 
At the same time it is desirable to minimize the computational difficulties 
connected with the introduction of certain metric conditions associated with 
partial control points. A corresponding orientation of the local Cartesian 
system has been described in Chapter IV - (D), page 35. 
Formulas describing refraction corrections have been derived by several 
authors. In [8] & treatment is given for astronomical refraction including 
zenith distances > 90°. In [9] expressions for target points inside the 
atmosphere have been derived. References with respect to corresponding basic 
assumptions and formulas may be found in the recently published report, [10] . 
A summary of the problem of refraction in photogrammetry is found in [11] e 
M, the meteorological constants referred to in formula (89) are either 
obtained in accordance with some model of the atmosphere or are determined 
  
directly from independent meteorological measurements, or obtained indirectly 
by additional elevation angle measurements with respect to known points; a 
i procedure which under certain conditions can be made a part of the actual 
| photogrammetric triangulation measurements. If aerial analytical triangulation 
| demands specific meteorological parameters in the area of the flight, it is 
feasible to eject periodically from the airplane probes for measuring densities 
I and/or temperatures; the corresponding results being recorded by radio link in 
the airplane. 
60