Optical Turbulence Profile in Marine Environment with Artificial Neural Network Model

Optical turbulence strongly affects different types of optoelectronic and adaptive optics systems. Systematic direct measurements optical profiles [Cn2(h)] are lacking for many climates seasons, particularly in marine environments, because it is impractical expensive to deploy instrumentation. Here, a backpropagation neural network optimized using genetic algorithm (GA-BP) developed estimate atmospheric environments which validated against corresponding profile datasets from field campaign balloon-borne microthermal at the Haikou environment site. Overall, trend magnitude GA-BP model agree. The generally superior those obtained by BP physically-based (HMNSP99) models. Several statistical operators were used quantify performance on reconstructing environments. characterization vertical distributions main integral parameters derived presented. median Fried parameter, isoplanatic angle, coherence time 9.94 cm, 0.69?, 2.85 ms, respectively, providing independent proposed approach exhibits potential implementation ground-based applications