The telescope performs a scan at a constant rate. The observer could select from three discrete rates — slow, medium, and fast (approximately 2.6, 6.5, and 17.0 arcseconds per second). A scan leg is defined to be a single scan pass in one direction (forward or reverse). A scan map observation always starts in the ''forward'' direction (in the +Y direction in the Spitzer focal plane). The scan length is defined to be the distance along the scan track for which coverage is obtained with all three MIPS arrays. Scan lengths can be as short as 0.5° or as long as 6° (depending on the scan rate). There are regions of overscan at each end that have coverage by some of the arrays but not all. During the full-coverage portion of the observation, the telescope position should be within at least 1.3´´ (rms) of the expected position along the scan track. At the end of a scan, the spacecraft pauses, then begins a new scan in the opposite direction to the first, with the same scan rate and either along the same track or offset by anywhere from 5´´ to 5´ in the cross-scan direction, as selected by the observer.
For typical scan maps, scan legs can be assumed to be linear. Very long scans depart slightly from linear tracks. This curvature is determined by the offset between the arrays and the telescope centerline (X-axis), and any misalignment between the scan mirror rotation axis and the spacecraft Z-axis (Solar direction). Also, the scanned region on the sky is in general not exactly 90º from the line to the Sun, but can be anywhere within a 40º wide annulus .The net effect of these terms on the scan leg geometry is that the dominant direction of scan motion is a great circle centered on the spacecraft Z-axis. The motion is made such that the rotation is around a point 90º from the telescope boresight, and within 2º of the great circle centered on Spitzer and going through the center of the Sun. The pointing of the boresight is such that the scan leg centerline is positioned through the array centers. Perturbations that amount to small circle deviations are required to account for the focal plane layout and the actual scan mirror rotation axis relative to the Z-axis. The resultant curvature in the scan leg is <1% over a 6º scan. The relative pointing accuracy between two consecutive scan legs is expected to be ultimately better than 3´´ (radial) for at least 4000 seconds time.
During the course of an AOR, the Spitzer focal plane orientation on the sky is fixed. Therefore, there is no rotation of individual scan paths, nor is there any relative rotation between scan legs obtained within a single AOR. This holds in spite of the orbital motion of Spitzer during the period of the AOR. Scans obtained in separate AORs likely will have different orientations. The magnitude of the rotation depends on ecliptic latitude and time between AORs. To minimize such variations in scan orientation, observations may have to be constrained such that they are done within a short period of time, particularly when the targets are far away from the ecliptic. Please see Spitzer Telescope Manual for more information about spacecraft alignments, axes, and constraints.