Effects of decoherence and propagation in remote detection of molecules using CARS

The backscattered signal of Coherent Anti-Stokes Raman Spectroscopy (CARS) is theoretically analyzed for remote detection of molecules in the atmosphere. An optimization of the CARS signal is done by maximizing vibrational coherence using chirped control pulses. The factors affecting the intensity of the output signal are investigated such as decoherence and multiple collisions with the target molecules upon propagation. The Liouville von Neumann equations with the relaxation terms of spontaneous decay and collisional dephasing are used. A numerical analysis of the dependence of the molecular state populations and coherence on the peak Rabi frequency and other field parameters is performed demonstrating adiabatic regime of light-matter interaction and mitigation of decoherence. For 100fs incoming pulses, a significant amplification of the anti-Stokes signal is demonstrated, up to 3 orders of magnitude after multiple scattering events. A machine learning technique is implemented to extract an analytical phase of scattered fields from their numerical values.