Implications of MODIS bow-tie distortion on aerosol optical depth retrievals, and techniques for mitigation

[摘要] The scan geometry of the Moderate Resolution Imaging Spectroradiometer(MODIS) sensors, combined with the Earth's curvature, results in a pixel shape distortion known as the "bow-tieeffect". Specifically, sensor pixels near the edge of the swath areelongated along-track and across-track compared to pixels near the centre ofthe swath, resulting in an increase of pixel area by up to a factor of ∼ 9 and, additionally, the overlap of pixels acquired from consecutive scans. The DeepBlue and Dark Target aerosol optical depth (AOD) retrieval algorithmsaggregate sensor pixels and provide level 2 (L2) AOD at a nominal horizontalpixel size of 10 km, but the bow-tie distortion means that they alsosuffer from this size increase and overlap. This means that the spatialcharacteristics of the data vary as a function of satellite viewing zenithangle (VZA) and, for VZA > 30°, corresponding to approximately50 % of the data, are areally enlarged by a factor of 50 %or more compared to this nominal pixel area and are not spatiallyindependent of each other. This has implications for retrieval uncertaintyand aggregated statistics, causing a narrowing of AOD distributions near theedge of the swath, as well as for data comparability from the application ofsimilar algorithms to sensors without this level of bow-tie distortion.Additionally, the pixel overlap is not obvious to users of the L2 aerosolproducts because only pixel centres, not boundaries, are provided within theL2 products. A two-step procedure is proposed to mitigate the effects of thisdistortion on the MODIS aerosol products. The first (simple) step involveschanging the order in which pixels are aggregated in L2 processing to reflectgeographical location rather than scan order, which removes the bulk of theoverlap between L2 pixels and slows the rate of growth ofL2 pixel size vs. VZA. This can be achieved without significantchanges to existing MODIS processing algorithms. The second step involvesadditionally changing the number of sensor pixels aggregated across-track asa function of VZA, which preserves L2 pixel size at around10 km × 10 km across the whole swath but wouldrequire algorithmic quality assurance tests to be re-evaluated. Both of thesesteps also improve the extent to which the pixel locations a user would inferfrom the L2 data products represent the actual spatial extent of theL2 pixels.

[学科分类] 几何与拓扑