Radiative transfer modeling and experimental validation of the multiple-scattering lidar signal propagation within...

The multiple-scattering return signals of cloud-detecting lidar contain a large amount of cloud characteristic information. Establishing an effective cloud-detecting lidar simulator is crucial for calibrating the accuracy of lidar data and developing inversion methods for cloud optical properties, as well as their macro- and micro-parameters. In this paper, the propagation of a collimated Gaussian laser beam in an inhomogeneous water cloud is first simulated based on the theory of atmospheric radiative transfer using Monte Carlo (MC) and semi-analytical MC methods. In the MC method, the photon state of the laser from the transmitter to the receiver is counted by tracking the scattering path of a large number of random photons. Conversely, the semi-analytical MC method relies on an analytical estimation of the scattered photon probability by the remote receiver. In particular, the vertical optical profile parameter for the water cloud input is no longer fixed at a constant value in both transport models, and the resolution of the vertical profile may also vary based on specific circumstances. Second, the effects of the optical properties and the receiving field of view (FOV) of water clouds on the results of lidar return signals are investigated using semi-analytical MC simulations in this paper. Finally, in the process of establishing the analytical model for the multiple-scattering return signals of water cloud lidar, relevant assumptions, such as the quasi-single-scattering small-angle approximation (QSA), are introduced. Based on this scattering principle, a multiple-scattering Raman lidar has been built at the Lund National Meteorological Station in the Liupan Mountains of Ningxia Province. We compared the semi-analytical MC simulation results, serving as a standard model, with measurements obtained from the lidar operating at the station. The absolute errors of the normalized simulated signals were within 0.18, demonstrating the accuracy of the analytical model and its associated assumptions. This finding suggests that all of the aforementioned methods are valid and applicable to inhomogeneous water clouds.