Langley Calibration Analysis of Solar Spectroradiometric Measurements: Spectral Aerosol Optical Thickness Retrievals

Aerosol optical thickness (τaer) is a fundamental parameter for analyzing aerosol loading andassociated radiative effects. Theτaercan constrain many inversion algorithms using passive/active sensormeasurements to retrieve other aerosol properties and/or the abundance of trace gases. In the next wave ofspectroradiometric observations from geostationary platforms, we envision that a strategically distributednetwork of robust, well-calibrated ground-based spectroradiometers will comprehensively complementspaceborne measurements in spectral and temporal domains. Spectralτaercan be accurately obtainedfrom direct-Sun measurements based on the Langley calibration method, which allows for the analysis ofdistinct spectral features of the calibration results. In this study, we present a spectralτaerretrieval algorithmfor an in-house developed,field deployable spectroradiometer instrument covering wavelengths fromultraviolet to near-infrared (UV-Vis-NIR). The spectral total optical thickness obtained from the Langleycalibration method is partitioned into molecular and particulate components by utilizing a least squaresmethod. The resulting high temporal-resolutionτaerand Ångström Exponent can be used effectively forcloud screening. The new algorithm was applied to month-long measurements acquired from the rooftopat National Aeronautics and Space Administration Goddard Space Flight Center’s Building 33. The retrievedτaerdemonstrated excellent agreement with those from well-calibrated Aerosol Robotic Network Sunphotometers at all overlapping wavelengths (correlation coefficients higher than 0.98). In addition, empiricalstray light corrections considerably improvedτaerretrievals at short wavelengths in the UV. The continuousspectrum ofτaerfrom UV-Vis-NIR spectroradiometers is expected to provide more informative constraints forretrieval of additional aerosol properties such as refractive indices, size, and bulk vertical distribution.