By uncorrelated I mean not mutually or reciprocally related. Sounds sum as vectors. When two sounds are of the same frequency and phase, they sum at any instant in time to a level that is twice as great in magnitude (if they are at the same level), i.e., the resultant is 6 dB greater. Some will question that and say it should be 3 dB, so I should remind them that we are talking about sound pressure levels, that is potentials rather than power measures. Sounds that are dissimilar in frequency and phase sum randomly from one instant to the next, and the resultant can only be quantified as an RMS sum over some finite interval. The RMS sum turns out to be 3dB above the component sounds. Examples of sounds of identical frequency and phase summing (said to be a coherent sum) are a monophonic sound played through your stereo. If your speakers are properly phased and balanced you hear the combined sound 6 dB louder apparently at a point midway between the speakers. If the speakers are miswired to be out of phase, the sounds are reciprocally related - still identical in frequency but with peaks and troughs in the waveforms opposite each other. In that case the vector sum of the sounds is -6dB relative to either, and you have a sound 'hole' in the middle. Similar relationships exist between prop blades, which is why prop noise is directional, peaking 10-20 degrees behind the plane of the prop disk. Prop and intake/exhaust sounds are related in frequency, at harmonics if not always significantly at fundamentals. The exhaust sound phase can be modified with reactive elements, and this leads to a possibility of cancellation between prop and exhaust emissions - but that's a whole other topic. Just a hint - ever notice a model with a tuned pipe that seemed much quieter than you expected? A tuned pipe has large capacitive reactance (compliance in acoustic terms)...........

I'm not familiar with a club field that mandates a racetrack pattern, nor with any AMA rule that does. That notwithstanding, I think you should consider 'persistence' as it relates to noise annoyance. It has been well established by research in the field that how often and how long a noise emission lasts is an important parameter in any measure of its propensity to annoy people. That is why the measure 'equivalent loudness' Leq is often used to characterize noise, in civic ordinances for example. Rather than instantaneous levels, average levels over some specified time period, often one hour, are used. The instantaneous levels could well be greater at some receptor's location under the situation you describe, but I think for practical purposes what counts is the level emitted at the center of the pattern(s) flown, however they meander.

Abel