SUMMARY

    Historically, the choice of formulae for model rocketry altitude 
data reduction has been limited to linear closed-form equations.  Using 
simple trigonometric functions, these equations require only a pocket 
calculator or, if necessary, could be derived by hand using sine and 
cosine tables. In recent years the widespread availability of 
programmable calculators and portable computers has automated these 
calculations.  In fact, the typical portable computer has more than 
enough computing power to perform vastly more complex methods of data 
reduction beyond the static closed-form algorithms. 
    The main objective of this project was to implement a computer 
software program for a more accurate method of determining a model 
rocket's likely altitude given two-station theodolite angles. The 
mathematical basis of the new method has a strong similarity to the 
closed form Geodesic method, but adds weighting for real-world effects 
of optically tracking a model rocket to ejection. The grouping of errors 
into a "region of uncertainty" is approximated by a hemispherical shape 
which is located using an iterative software technique, hence the name 
"Hemispherical Iteration Tracking". 
    The software was tested for various regions in 3D space relative to 
the trackers and the convergence of the HIT method was compared to the 
Geodesic method with favorable results.  Additionally, a set of data 
from a particularly problematic NAR-sanctioned regional was used to 
compare the new method with the Geodesic equations; the results showed 
that the closure rate would have improved from only 31% of tracked 
flights closed to over 88% closed.
    NAR competitors would benefit from this new data reduction method by 
improving the closure rate for altitude events, offering a more even 
playing field when existing algorithms fail to close reliably under 
certain conditions.  Further statistical analysis would be needed before 
approving the new method for NAR-sanctioned competition and record 
attempts.