Direct retrieval of sulfur dioxide amount and altitude from spaceborne hyperspectral UV measurements: Theory and application

We describe the physical processes by which a vertically localized absorber perturbs the top‐of‐atmosphere solar backscattered ultraviolet (UV) radiance. The distinct spectral responses to perturbations of an absorber in its column amount and layer altitude provide the basis for a practical satellite retrieval technique, the Extended Iterative Spectral Fitting (EISF) algorithm, for the simultaneous retrieval of these quantities of a SO 2 plume. In addition, the EISF retrieval provides an improved UV aerosol index for quantifying the spectral contrast of apparent scene reflectance at the bottom of atmosphere bounded by the surface and/or cloud; hence it can be used for detection of the presence or absence of UV absorbing aerosols. We study the performance and characterize the uncertainties of the EISF algorithm using synthetic backscattered UV radiances, retrievals from which can be compared with those used in the simulation. Our findings indicate that the presence of aerosols (both absorbing and nonabsorbing) does not cause large errors in EISF retrievals under most observing conditions when they are located below the SO 2 plume. The EISF retrievals assuming a homogeneous field of view can provide accurate column amounts for inhomogeneous scenes, but they always underestimate the plume altitudes. The EISF algorithm reduces systematic errors present in existing linear retrieval algorithms that use prescribed SO 2 plume heights. Applying the EISF algorithm to Ozone Monitoring Instrument satellite observations of the recent Kasatochi volcanic eruption, we demonstrate the successful retrieval of effective plume altitude of volcanic SO 2 , and we also show the improvement in accuracy in the corresponding SO 2 columns.