A54B-05:
A new paradigm for constraining PM2.5 speciation by combining multiangular and polarimetric remote sensing with chemical transport model information
Friday, 19 December 2014: 5:00 PM
Olga Kalashnikova1, Feng Xu2, Cui Ge3, Jun Wang3, Michael J Garay4 and David J Diner5, (1)NASA Jet Propulsion Laboratory, Pasadena, CA, United States, (2)JPL/UCLA Joint Institute for Regional Earth System Science and Engineering, Pasadena, CA, United States, (3)University of Nebraska Lincoln, Lincoln, NE, United States, (4)Jet Propulsion Laboratory, Pasadena, CA, United States, (5)JPL, Pasadena, CA, United States
Abstract:
Exposure to ambient particulate matter (PM) has been consistently linked to cardiovascular and respiratory health effects. Although PM is currently monitored by a network of surface stations, these are too sparsely distributed to provide the level of spatial detail needed to link different aerosol species to given health effects, and expansion to denser coverage is impractical and cost prohibitive. We present a methodology for combining Chemical Transport Model (CTM) aerosol type information and multiangular spectropolarimetric data to establish the signature of specific aerosol types in top-of-atmosphere measurements, and relate it to speciated surface PM2.5 loadings. In particular, we employ the WRF-Chem model run at the University of Nebraska, and remote sensing data from the Airborne Multiangle SpectroPolarimetric Imager (AirMSPI) to explore the feasibility of this approach. We demonstrate that the CTM does well in predicting the types of aerosols present at a given location and time, however large uncertainties currently exist in CTM estimates of the concentration of the various aerosol species (e.g., black carbon, sulfate, dust, etc.) leading to large uncertainties to model-derived speciated PM 2.5. In order to constrain CTM aerosol surface concentrations we use AirMSPI UV-VIS-NIR observations of intensity, and blue, red, and NIR observations of the Q and U Stokes parameters. We select specific scenes observed by AirMSPI and use WRF-Chem to generate an initial distribution of aerosol composition. The relevant optical properties for each aerosol species are used to calculate aerosol light scattering information. This is then used in a vector (polarized) 1-D radiative transfer model to determine at-instrument Stokes parameters for the specific AirMSPI viewing geometries. As a first step, a match is sought between the CTM-predicted radiances and the AirMSPI observations. Then, the total aerosol optical depth and fractions of various aerosol species are modified via optimization to produce a better match to the observations, and converted to PM2.5 speciated loadings using CTM aerosol vertical profiles. Finally, the results are compared to available ground-based and in situ data to validate this approach.
