Interleaving and repeat coding are techniques that provide immunity to channel effects for optical communications links. Interleaving provides robustness at the physical layer of the network stack, at the expense of additional latency. The temporal diversity ratio, L, is defined as the number of statistically-independent fade events experienced by an individual code-word in an interleaved framing structure. L characterizes an interleaver's ability to whiten a channel so forward-error correction codes for mitigating additive white Gaussian noise function effectively. For large L's, an interleaver can tolerate a wide duration of fade dropouts with high latency, while smaller L's improve latency yet suffer more fade-induced power penalties. Repeat coding is another technique that can enhance a channel's resilience to atmospheric fading by sending identical copies of frames. Repeatcoding techniques allow a link to operate at lower signal-to-noise ratios, while still maintaining a fixed slot clock to simplify receiver design. We define Q as the frame replication factor for a repeat-coded waveform.We investigate interleaver performance in different regimes of both L and Q. We develop a model to supplement interleaver dynamics with repeat coding under atmospheric fading, and investigate regimes with different values of fading strength, interleaver diversity L, and repeat-coding Q. An experimental optical modem testbed with a flexible configuration is used to validate the model. We show good agreement between the model and experimental laboratory tests over a wide range of cases studied.