Modeling Multiple Scattering and Absorption for a Differential Absorption LIDAR System

摘要

The Digital Image and Remote Sensing Image Generation (DIRSIG) model has been developed and utilized to support research at the Rochester Institute of Technology (RIT) for over a decade. The model is an established, first- principles-based scene simulation tool that has been focused on passive multi- and hyper-spectral sensing from the visible to long wave infrared (0.4 to 14 micrometers). Leveraging photon mapping techniques utilized by the computer graphics community, a first-principles-based elastic Light Detection and Ranging (LIDAR) model was incorporated into the passive radiometry framework so that the model calculates arbitrary, time-gated photon counts at the sensor for atmospheric, topographic, and backscattered returns. The active LIDAR module handles a wide variety of complicated scene geometries, a diverse set of surface and participating media optical characteristics, multiple bounce and multiple scattering effects, and a flexible suite of sensor models. This robust modeling environment allows the researcher to evaluate sensor design trades for topographic systems and the impact that scattering constituents (e.g. water vapor, dust, sediment, soot, etc.) may have on a Differential Absorption LIDAR (DIAL) system's ability to detect and quantify constituents of interest within volumes including water and atmospheric plumes. The interest in modeling DIAL sensor engagements involving participating media such as gaseous plumes presented significant challenges that were overcome using the photon mapping paradigm. Intuitively, researchers suspect that multiple scattering effects from additional constituents as simple as water vapor or soot could impact a DIAL sensor's ability to detect and quantify effluents of interest within a participating medium. Traditional techniques, however, are not conducive to modeling the multiple scattering and absorption within a non-homogenous finite volume, such as a plume.