Location
Off Campus : via Webex
Date & Time
August 10, 2020, 4:00 pm – 6:00 pm
Description
ADVISOR: Dr. J. Vanderlei Martins
TITLE: Vertically Resolved Aerosol Retrievals Using Combined LIDAR/HARP-Polarimeter Method
ABSTRACT: Aerosols remain one of the most substantial unknowns in climate science, with their radiative forcing effect on the planet potentially being either positive or negative, depending on their composition and location in the atmosphere. Particularly influential is their height and that height in relation to clouds whose properties change due to aerosol influence. In light of this, remote sensing techniques and platforms have been and are being developed to help atmospheric scientists study the real properties and distributions of aerosols around the globe. Passive sensors take pictures of reflectance and emissions, while active sensors literally shine a light on the atmosphere and capture the results. Aerosol properties can be “retrieved” from both techniques by inverting the problem of radiative transfer, a first-order integral-differential equation which is not, in general, analytically invertible.
In light of this, retrievals are typically done through numerical means, either via interpolation of long used look-up tables containing pre-tabulated results for a variety of inputs, or through the careful fitting of measurements to the output of the radiative transfer solution via numerical regression. The latter has come into an era of its own and allows the retrieval of far more informationally dense retrievals of passive polarimetric imagery than the look-up table allowed. The HyperAngular Rainbow Polarimeter (HARP) instruments, designed and built at UMBC, stand among the leading polarimeters capable of informing such a retrieval. What all passive sensors lack though is a simple way to vertically resolve the aerosol plumes they measure, relying instead on column integrated totals. Active LIDAR, on the other hand, can quite simply isolate vertical layers, but its narrow swath will never allow full global coverage like a passive imager can and its dependence on backscatter only lacks sensitivity to particle shape. Both these limitations are fundamental problems of specialization in instrument design which can be resolved by taking simultaneous measurements from these very different platforms and performing the regression on their combined data.
To perform a combined LIDAR+HARP Polarimeter retrieval, we must apply the methods of regression and radiative transfer that have seen widespread use, such as in the Generalized Retrieval of Aerosol and Surface Properties (GRASP). To utilize the numerical regression scheme though we must identify constraints on the regression, sensitivity of the retrieval to the measurements, and validate the results. To achieve all these, I propose to: 1; Develop a HARP polarimetric error model fully accounting for alignment and calibration, and 2; From the HARP error model, develop a LIDAR/Polarimeter Observational System Simulation Experiment (OSSE) which would seek to create virtual aerosols that can be retrieved the same as real measurements. This would provide a ground truth with which to validate a joint retrieval framework. Real world data for these retrievals also exists from the airborne campaigns of AirHARP which flew on the same the newest High Spectral Resolution Lidar (HSRL-2) from NASA Langely. These modeling will also inform the next HARP iteration, HARP2 to be launched in 2022 aboard the much-anticipated Plankton, Aerosol, Cloud, and ocean Ecosystems (PACE) mission and whatever may follow.
Proposal will be held using Webex.
TITLE: Vertically Resolved Aerosol Retrievals Using Combined LIDAR/HARP-Polarimeter Method
ABSTRACT: Aerosols remain one of the most substantial unknowns in climate science, with their radiative forcing effect on the planet potentially being either positive or negative, depending on their composition and location in the atmosphere. Particularly influential is their height and that height in relation to clouds whose properties change due to aerosol influence. In light of this, remote sensing techniques and platforms have been and are being developed to help atmospheric scientists study the real properties and distributions of aerosols around the globe. Passive sensors take pictures of reflectance and emissions, while active sensors literally shine a light on the atmosphere and capture the results. Aerosol properties can be “retrieved” from both techniques by inverting the problem of radiative transfer, a first-order integral-differential equation which is not, in general, analytically invertible.
In light of this, retrievals are typically done through numerical means, either via interpolation of long used look-up tables containing pre-tabulated results for a variety of inputs, or through the careful fitting of measurements to the output of the radiative transfer solution via numerical regression. The latter has come into an era of its own and allows the retrieval of far more informationally dense retrievals of passive polarimetric imagery than the look-up table allowed. The HyperAngular Rainbow Polarimeter (HARP) instruments, designed and built at UMBC, stand among the leading polarimeters capable of informing such a retrieval. What all passive sensors lack though is a simple way to vertically resolve the aerosol plumes they measure, relying instead on column integrated totals. Active LIDAR, on the other hand, can quite simply isolate vertical layers, but its narrow swath will never allow full global coverage like a passive imager can and its dependence on backscatter only lacks sensitivity to particle shape. Both these limitations are fundamental problems of specialization in instrument design which can be resolved by taking simultaneous measurements from these very different platforms and performing the regression on their combined data.
To perform a combined LIDAR+HARP Polarimeter retrieval, we must apply the methods of regression and radiative transfer that have seen widespread use, such as in the Generalized Retrieval of Aerosol and Surface Properties (GRASP). To utilize the numerical regression scheme though we must identify constraints on the regression, sensitivity of the retrieval to the measurements, and validate the results. To achieve all these, I propose to: 1; Develop a HARP polarimetric error model fully accounting for alignment and calibration, and 2; From the HARP error model, develop a LIDAR/Polarimeter Observational System Simulation Experiment (OSSE) which would seek to create virtual aerosols that can be retrieved the same as real measurements. This would provide a ground truth with which to validate a joint retrieval framework. Real world data for these retrievals also exists from the airborne campaigns of AirHARP which flew on the same the newest High Spectral Resolution Lidar (HSRL-2) from NASA Langely. These modeling will also inform the next HARP iteration, HARP2 to be launched in 2022 aboard the much-anticipated Plankton, Aerosol, Cloud, and ocean Ecosystems (PACE) mission and whatever may follow.
Proposal will be held using Webex.
