The surface integral approach to Radarclinometry

Because radarclinometry is fundamentally describable in terms of a nonlinear, first-order, partial differential equation, one expects that it can, in principle, be carried out by direct deterministic integration beginning at a given threshold profile along the azimuthal coordinate. Such a boundary condition could be provided by the altimetry profile obtained on a preceding or succeeding orbital revolution of the radar-bearing spacecraft. Notwithstanding the mismatched resolutions of the radar altimeter and the radar imaging system as planned for the Megallan mission to Venus, there are fundamental considerations, not involving system noise, that influence the possibility of success of this approach. From the topographic map of the Lake Champlain West quadrangle in the Adirondack Mountains of the U.S., a radar image is synthesized. Radarclinometry, in surface integral form, recaptures the topographic map when the applicable radar reflectance function is weakly variable over the range of application, but it diverges beyond a certain point for nominally variable reflectance functions. The effect can be understood by using results from the "shape-from-shading" literature. (This literature is produced by a group within the artificial intelligence community who have been independently attacking, for all practical purposes, photoclinometry, except that they have not given primacy to images of terrain.) The ubiquity of the instability suggests that the value of the surface integral approach is much in doubt. ?? 1988 Kluwer Academic Publishers.