Spectral Differential Imaging (SDI) and Angular Differential Imaging (ADI) are the base modes of the NICI instrument.  Whether SDI is done at discrete roll angles or combined with ADI is the choice of the observer and is not necessarily a straightforward decision.

Spectral Differential Imaging

The SDI technique seeks to improve the separation of real objects from speckles. SDI achieves this by exploiting a spectral feature in the desired target. The common implementation utilizes the 1.6 um methane absorption feature robustly observed in substellar objects cooler than 1400 K.

Images are taken simultaneously both within and outside the methane absorption feature.  Due to the simultaneity of the observations, the star and the coherent speckle pattern are largely identical in both filters, while any faint companion with that absorption feature is brighter in one filter than in the other. Subtracting the two images thus removes the starlight and speckle patterns while a methane containing companion remains in the image.

To confirm a detection and to partly reduce instrumental speckles SDI can be used at two discrete rotator angles for each object.

A real object on the sky will rotate with a change of rotator angle; however, speckles caused inside the instrument will not. By subtracting data taken at multiple roll angles, speckles are further attenuated and any real object in the frame will display the corresponding roll angle jump.

The disadvantage of using discrete roll angles versus ADI is that the pupil moves with respect to the image. The moving spider arms (support of the secondary mirror) cause a changing diffraction pattern. Therefore, the instrumental speckles are not fully stable at different roll angles and PSF subtraction is less accurate.

Angular Differential Imaging

To apply the ADI technique the rotator of an altitude-azimuth telescope is deliberately deactivated (fixed position) which causes the telescope optics to rotate on the sky. Thus, the field of view in a sequence of images rotates and a real companion will move on the image according to the change of parallactic angle.

For most of the used total exposure times (e.g. larger than 10 sec), speckles caused by the turbulent atmosphere average out. With the pupil not moving the instrumental speckle pattern stays stable and in that sense allows for an ideal PSF subtraction.

The disadvantage is that the field rotation depends on the target position on the sky and the PSF will smear according to the exposure time and distance to the pivot point in the image. Nevertheless, it turns out that for many targets of interest, the field rotation has suitable rates with a negligible PSF smearing in the region of interest (i.e. within the inner 2" of the central star).

As shown in the figure below, typical field rotation rates are between 1 to 2 degrees per minute. When the target declination is close to the Cerro Pachon latitude (-30.2 deg) an accurate timing window is needed. Otherwise field rotation will be close to zero for most of the time and explode while passing the meridian.

With NICI the ADI and SDI techniques can be combined, thus providing an even greater degree of speckle suppression.

Figure: The figure on the top shows the rotation of the Field of View for Hour Angles ranging from 180 minutes before to 180 minutes after passing the meridian (ADI mode, rotator fixed). The figure below displays its derivative. For example, it shows that for all targets with -75 deg > DEC > -55 deg or -5 deg < DEC <  15 deg the field rotation rates are between 1 to 2 degrees per minute.

References:  Biller et al., 2008, SPIE; Artigau et al., 2008, SPIE.