Developing and Reporting an Uncertainty Metric for Satellite Aerosol Retrievals and Comparisons with Ground-Based Sensors

Over the past few years, the algorithm for retrieving atmospheric aerosols from the Multi-angle Imaging SpectroRadiometer (MISR) instrument on NASA’s Terra satellite has undergone extensive revision and testing.  The outcome of this effort is an aerosol product with higher spatial resolution (4.4 km) in comparison with the previous 17.6 km spatial resolution product.  Additionally, an approach was developed to report pixel-level uncertainties in aerosol optical depth (AOD), the fundamental retrieval quantity that relates observations made by the MISR instrument to the amount of aerosols in the atmospheric column.  This uncertainty, in particular, has been frequently requested by consumers of such satellite products.  However, it is unclear whether or not the reported field will necessarily meet the needs of the community, but providing it on a global basis is an important first step that will hopefully stimulate discussion.

We will describe the development and testing of the MISR AOD uncertainty field, which corresponds to the uncertainty in the retrieval itself, as opposed to some comparison with an external “truth” dataset.  We will also discuss the spatial and temporal variability of the ground-based datasets used for intercomparison and validation and how these affect the assessment of the performance of the MISR aerosol product and the associated uncertainty.