|
A dual-band imaging system with variable aperture diameter was constructed and horizontal and vertical atmospheric tilt components were measured on a 1-km near-the-ground horizontal path using discrete and extended visible and JR sources. The spatial and temporal tilt statistics were estimated from the recorded data. Tilt structure function, which also characterizes v ariance of the p ointing error caused by anisoplanatism of t he track point to the aim point in the 1 aser projection system, for small angular separation decreases inverse proportionally to the aperture diameter D1 . The tilt structure function is insensitive to sensor vibration. For a point ahead angle of 0.45 mrad the daytime rms pointing enor caused by tilt anisoplanatism is 12 prad for D= 6 cm, and it is 5 prad for D= 40 cm. The tilt power spectral density agrees well with theory. Jt has the "-2/3" power slope, and the ratio of the two knee frequencies is equal to the inverse ratio of the aperture diameters. The tilt temporal conelation increases with the aperture diameter. The temporal conelation scale is 0.25 sec for D=6 cm and it is 1 sec for D= 40 cm. The C measurements made with discrete JR sources and an optical imager agree well with the measurements made with a scintillometer. The structure function for the lateral (Y) tilt exceeds that for the longitudinal (X) tilt, which is inconsistent with the theoretical prediction. We believe that heat-induced turbulence from the JR sources and a wind component parallel to the optical path degraded the measurements of the vertical tilt. Three mitigation techniques were considered including an increase of the aperture diameter, integration of the image edge over the edge angular extent, and averaging of multiple frames. A multi-frame averaging technique is known to be efficient for mitigation of the effects of turbulence induced scintillation and laser speckle. We found that by averaging multiple image frames one can mitigate the effects of tilt anisoplanatism as well. We also found that the edge response for a multi frame averaged image and a single frame image is the same. This allows us to conclude that a multi frame averaging technique for an extended object does not affect the system angular resolution.
|
