Information about global distributions of aerosol optical thickness (AOT) and size categorization is necessary to quantify the aerosol radiative forcing as well as a method to monitor air-quality such as fine particulate matter (PM). Efforts to provide such aerosol optical properties on a global scale clearly require satellite retrievals and therefore validation of satellite products is a clear priority. The development of suitable radiometer networks is central to this effort. Modern robotic solar instruments for retrievals of aerosols optical depth and other microphysical parameter retrievals, such as the CIMEL Sky Scanning Radiometer [Holben, 1998] as part of the Aeronet Network, are crucial to this effort but are unfortunately quite sparse and expensive and efforts to develop and utilize simpler portable instruments are highly desirable One attractive possibility is the deployment of Multi-Filter Rotating Shadow-band Radiometer (MFRSR) in a network [Alexandrov, 2002] both for validation efforts as well as monitoring extended megacities such as the NYC area.

Furthermore, the developments of these portable networks in urban areas are particularly crucial for the following reasons: (1) The retrieval of aerosol properties over land areas and particularly urban surfaces is far more challenging than over oceans due to the bright and complex surfaces in urban areas. Cutting Edge algorithms such as the aerosol retrieval from the Advanced Polari-metric Scanner (APS) [Waquet, 2009] from current and future platforms are in need for validation and clearly would benefit from instruments that can be moved to strategic locations under the satellite track. In addition, for climate studies, separation of aerosols into fine/coarse mode constituents [Mischenko, 2004] as well as identifying and quantifying absorbing aerosols is critical. These outputs are in fact a major focus of APS (or combined APS-MODIS) and pulling these parameters out from the MFRSR processing is critical in maximizing the value. (2) In addition, urban areas have more spatial diversity in aerosols and would greatly benefit from closely spaced measurements that can probe local structure and anomalies.

In this thesis, our goal is to test and implement a newly developed retrieval algorithm developed at NASA GISS [Alexandrov et al. 2006] for processing MFRSR data. In particular, we show that this algorithm significantly improves optical depth time series measurements in comparison to the most currently used Langley regression method calibration. We report our deployment of the MFRSR network over the NYC metropolitan area and describe intercomparison measurements between AOD measurements both inside and outside NYC in an effort to explore local aerosol production. Finally, the MFRSR network shall provide us with the spatial and temporal resolution needed to validate satellite data.

The thesis contents are as follows. In section 1, motivation of the importance of aerosols for both climate impacts and human health are briefly given as well as a primer on what optical quantities we are measuring to detect aerosols. In section 2, we briefly discuss the different ways to measure aerosols including satellite and ground remote sensing (and in-situ) instruments with a particular focus on establishing networks that can validate complex aerosol satellite algorithms. In section 3, we present our validation of the MFRSR algorithm of NASA GISS [Alexandrov et al. 2002, 2005, 2007, 2008] against CIMEL measurements to ensure that we are handling the processing stream properly. In section 4, we present our current and future network topology, processing, networking etc as well as present preliminary multisensory matchups for total AOD, fine and coarse mode separation and water vapor with particular focus on correlations and discrepancies. In section 5, we illustrate the value of the network through satellite validations for both MODIS and GOES aerosol retrievals. In section 6, we present our efforts to date in trying to pull out single scattering albedo and future prospects in moving the network deployment and algorithms forward. (Abstract shortened by UMI.)