Atmospheric wind speed profiling fi om an Earth-orbiting satellite with an active lidar probe is considered. "Direct detection" of the Doppler shift, i.e., with optical interferometry, can employ the ultraviolet, and thus can be based on Rayleigh (molecular) backscatter, permitting measurements without dependence on atmospheric aerosol content, Two direct detection methods,"edge technique" and "fringe imaging", are compared here for measurement sensitivity in the signal shot noise limit. There is no significant difference; both are capable, at best, of measurement statistical uncertainty about 2 to 4 times the Cramer-Rao limit. of a perfect, lossless receiver. With this result, and reasonable parameters for a satellite instrument, one can determine that the minimum laser power - telescope aperture product needed to make measurements with sufficient accuracy to be useful in atmospheric models is on the order of a few hundred W-m(2). For practical lasers and satellites, this implies an optical receiver of 2 to 4 meters diameter. Lightweight, deployable optical collectors, currently being developed, may economically provide these large apertures, with optical quality adequate for noncoherent lidar. The large aperture collector requires a Doppler analysis interferometer with high etendue. With practical values for the lidar receiver field of view, the Fabry-Perot etalon for the fringe imaging technique must be about 50-135 mm in working aperture, while the edge technique would require 200-540 mm aperture. Since practical etalons are limited to about 160 mm working aperture, the edge technique appears to be unsuitable for the spaceflight Doppler wind lidar. A conceptual design of a spaceflight Doppler wind lidar system is shown, assuming hardware parameters that appear to be feasible within the near-term state of the art. The very large Doppler shift due to spacecraft motion can be dealt with by tracking the signal through several orders of the Fabry-Perot etalon. The background noise introduced by solar backscatter from a high-albedo background can be adequately suppressed with a dual tandem etalon preceding the Doppler analysis etalon, if fabrication techniques for spaceflight etalons can be moderately improved.