Future 3D-Wind Measurements from Multi-Satellite Observations: A Demonstration with MISR and GOES-R

The GOES constellation has long been used to observe atmospheric motions by tracking the evolution of cloud and moisture patterns from geostationary orbit. A single GOES satellite can measure the horizontal displacement well over time but provides no direct height information. Height/pressure assignments for GOES Derived Motion Winds (DMWs) are typically inferred from IR brightness temperatures and modeled atmospheric profiles. The MISR instrument on the other hand offers the capability to determine pattern heights from first-principles observations of parallax as its cameras acquire multiple looks at cloud patterns along the orbital track of NASA's Terra spacecraft. MISR is well suited to pattern height determination but pattern motion along track is aliased with parallax. In this paper, we show the benefits of fusing the pattern tracking capabilities of GEO (e.g., GOES-R) and LEO (e.g., MISR) systems for the problem of 3D-Winds. The new approach relies on tracking an observed pattern across multiple looks from each system and between the GEO and LEO satellites. The 3D retrieval algorithm fits observed sub-pixel disparities to model pattern motion and height and explicitly accounting for non-simultaneity. The high accuracy Image Navigation and Registration (INR) of GOES-R is an important enabler of the multi-satellite stereo approach, allowing more comprehensive studies of atmospheric dynamic processes such as hurricanes, planetary boundary layer, and deep convection, with vertically-resolved wind fields. The new algorithm has capability to determine systematic errors in the geometry between the two satellites and compensate them through a bundle adjustment, to further improve accuracy. The retrieved 3D-Wind fields are evaluated against operational GOES DMWs and MISR cloud products. Demonstration of the LEO-GEO 3D-Winds algorithm in this study has an important implication for future global 3D wind observations that utilize GEO satellites and a low-cost cubesat constellation.