In answer to your question, yes there is a way. I even have a VB3 program that does it. The sensitivitiy to timing error is remarkably great. Also, the Cd you measure in a drop is apt to be in a rapidly declining region of the Cd/speed (or reynolds number if you prefer) graph. It will likely be higher than a Cd that accurately describes your flights.
(One of these days, I'll convert the drop program to an Excel spreadsheet. The solver can be invoked to do the backtracking. It still won't solve the speed problem)
I suggest you use a range of Cd's. For example, I use
0.5 - 0.9. If, for example, you are trying to determine optimal mass (very important for water rockets) you can1) Determine optimal mass at the extreme values.
2) Average the optimal masses you derived there; 3) Substitute the average mass at the extremes; and 4) Observe that the altitude penalty for the error is negligible.If your simulations are good, and if you have a pressure gage, then backtrack the Cd from observed altitudes. Would suggest that you use no ballast in Cd determination flights. Extra-light rockets will not go as high, but the flights will have a lot of drag influence, so your errors will be smaller. Use average values from many flights.
FWIW, near optimal mass, Cd has virtually no affect on
**cutoff** altitude and velocity.