Future Mission Concept for Operational Retrieval of Cloud-Top Heights and Cloud Motion Wind Vectors

To meet operational demands for vertically-resolved atmospheric winds, we propose a wide-swath multiangle stereo imaging system that could be easily integrated into future satellite platforms. The concept uses techniques proven with the multi-angle imaging spectroradiometer (MISR) instrument, currently flying in polar low-Earth orbit (LEO) on NASA's Terra satellite. MISR has demonstrated the feasibility of obtaining accurate stereophotogrammetric retrievals of cloud and aerosol plume heights and height-resolved cloud-tracked wind vectors. MISR obtains multiangle imagery with along-track pointing angles ranging from nadir to 70deg forward and backward of nadir and makes use of the several-minute time delay among a triplet of multiangle views to separate the effects of wind displacement from height-induced stereoscopic parallax. Careful attention to multiangle image co-registration (crucial for obtaining accurate wind retrievals) has led to wind estimates that exhibit negligible bias relative to reanalysis data. The root-mean-square uncertainty in speed is estimated to be in the 1-3 m/sec range, depending on whether the wind component is parallel or perpendicular to the flight direction and on the degree of quality control applied. The uncertainty in height assignment is on the order of 300 mum. Comparison of results from a forward-viewing angle triplet with independent retrievals from a backward-viewing triplet provides an important control on retrieval quality. A key attribute of the stereophotogrammetric approach to retrieving cloud heights (and winds) is that the technique is purely geometric. Accurate radiometric calibration is not required, thereby enabling significant simplification of a future instrument dedicated to stereo height and wind retrievals. Furthermore, the approach is insensitive to atmospheric temperature profiles and cloud emissivities, and provides data of consistent quality at all latitudes from polar orbit. It is also purely passive, so sensor lifetime is not a major issue. Since the technique relies on pattern matching, only a single mid-visible spectral band is required (a separate thermal infrared band, e.g., at 3.7 mum, would be needed in order to acquire measurements at night). Future improvements in computer capability will enable improved pattern-matching algorithms that could be implemented operationally. This simple, compact, and relatively inexpensive concept would assist numerical weather prediction models by retrieving cloud-motion vector winds in traditionally data-sparse areas (mid-oceans and high latitudes), would continue the unique data record begun with MISR by establishing long-term cloud-top height and cloud-motion wind monitoring with sensitivities capable of detecting atmospheric response to climate forcings, and would provide global measurements of the injection and transport heights of distinct aerosol plumes from wildfires, dust storms, and volcanic eruptions.