AUSTRALIAN NATIONAL UNIVERSITY System Design Note 4.11 Created: 30 June 2000 Last modified: 13 July 2000

STRATEGIES FOR TARGET ACQUISITION WITH NIFS

Peter J. McGregor

Research School of Astronomy and Astrophysics

Institute of Advanced Studies

Australian National University

Revision History

Revision No.	Author & Date	Approval & Date	Description
Revision 1	Peter J. McGregor 01 July 2000	John Hart 14 July 2000	Original document.

Contents

1 Purpose. 1

2 Applicable Documents. 2

3 Introduction. 2

4 NIFS Performance As An Acquisition Camera. 2

5 Is Collapsing Dispersed Images Sufficient?. 5

Appendix A: List of Figures. 5

1 Purpose

This document describes strategies for acquiring science targets with the Gemini Near-infrared Integral Field Spectrograph (NIFS) and tabulates associated performance predictions.

2 Applicable Documents

Document ID	Source	Title
SDN0003.01	RSAA	NIFS Operational Concept Definition Document
SDN0004.01	RSAA	NIFS Performance Model
SDN0005.12	RSAA	Atmospheric Dispersion Effects in NIFS

3 Introduction

The Gemini Near-infrared Integral Field Spectrograph (NIFS) will have a 3.0²´3.0² field-of-view. Consideration must therefore be given to how faint science targets will be acquired in order to maximize the observing efficiency of the instrument. These strategies are described in the NIFS Operational Concept Definition Document (OCDD; SDN0003.01) based on the detailed considerations presented here.

The Gemini telescopes will be used in both classically scheduled and queue scheduled modes. It is expected that observations will proceed with the minimum of observer intervention. Peripheral wavefront sensor (PWFS) guide stars will be selected prior to the observation. PWFS1 will acquire its guide star after the telescope is slewed to a new object. PWFS2 is not available when NIFS is used with ALTAIR and the AO fold mirror is deployed. Active correction of the primary mirror will begin once the PWFS1 guide star has been acquired. The NIFS science target should then be centered in the 3.0²´3.0² NIFS science field to an accuracy of ~ ±TBD² which includes the setting accuracy of PWFS1 and likely flexure through ALTAIR to NIFS. This will be sufficient for many science programs. However, unacceptable centering errors may result from a) differences between the optical coordinate system (DSS, USNO, HST GSC, etc.) and the infrared coordinate system (2MASS, DENIS, etc.), b) excessive flexure of NIFS/ALTAIR relative to PWFS1, and c) uncorrected differential refraction effects (Atmospheric Dispersion Effects in NIFS, SDN0005.12).

NIFS will use a duplicate of the NIRI On-Instrument Wavefront Sensor (OIWFS) to correct differential flexure between ALTAIR and NIFS. However, the setting accuracy of the OIWFS X-Y gimbal mirror is expected to limit its absolute pointing accuracy to ~ ±1². Consequently, the OIWFS cannot be used to improve on the absolute pointing defined via PWFS1.

Accurate target acquisition with NIFS will require the science field to be imaged through the NIFS science instrument. This can potentially be done in two ways; a) by collapsing a dispersed spectral image in the spectral direction, b) by recording an undispersed broad band image. Undispersed broad band images can be recorded by either rotating the grating to a mirror on the grating wheel, or by inserting a mirror in the collimated beam slightly in front of the grating Both possibilities for recording undispersed broad band images have their shortcomings. Moving the grating wheel demands extremely high levels of grating setting reproducibility in order not to compromise wavelength and flat fielding stability. Inserting a mirror in front of the grating requires an additional mechanism. This has budget and schedule implications, and it remains unclear whether sufficient space is available for such a mechanism.

This document addresses the performance aspects of collapsing dispersed spectral images and recording undispersed images in order to assess the design trades involved.

4 NIFS Performance As An Acquisition Camera

NIFS will achieve signal-to-noise ratios of ~ 10 per spectral pixel in a 0.1²´0.1² aperture in median seeing conditions in 1800 s on point sources with J = 18.4, H = 18.8, and K = 17.8 mag (NIFS Performance Model; SDN0004.01). Limiting magnitudes of J = 20.9, H = 21.3, and K = 20.3 mag are expected (~ 3s in 5 hr).

Short integration times of ~ 10-30 s will be required for efficient acquisition using the NIFS science instrument. The frame read time of the NIFS science detector is expected to be ~ 10 s so read noise reduction using multiple samples will not be possible on these timescales. The read noise of the HAWAII-2 detector in single-sample mode is expected to be ~ 15 e. Lower read noise of ~ 10 e may be possible in 100 s acquisition exposures.

The NIFS performance model has been used to estimate imaging performance based on both dispersed and undispersed data frames with a 10 s integration time and 15 e read noise. Short duration dispersed exposures will be dominated by read noise. Averaging 2048 spectral pixels leads to a factor of Ö2048 = 45.3 noise reduction to ~ 15/45.3 = 0.33 e. Short duration undispersed exposures will be dominated by background noise from airglow emission in the J and H bands and a combination of airglow emission and thermal emission principally from ALTAIR in the K band. This background noise prevents the full signal-to-noise ratio gain from imaging light on just one pixel from being realized in undispersed images. The undispersed background signal in a 10 s exposure will typically be ~ 2000 e per 0.04²´0.04² pixel in the H band. This produces a noise of ~ 45 e. The broad band signal detected per 0.04²´0.04² pixel in the undispersed case is ~ 2048/2 = 1024 times larger than in the dispersed case, so short duration undispersed images should achieve a signal-to-noise ratio ~ 1024*0.33/45 = 7.5 times larger than images based on dispersed data, or alternatively probe fainter by ~ 2.2 mag.

Masking OH airglow emission-lines in the dispersed images does not significantly improve sensitivity due to the dominance of read noise.

Dispersed acquisition exposures gain sensitivity in proportion to exposure time (because they are read noise limited) while undispersed acquisition exposures gain sensitivity only as the square root of exposure time (because they are background limited). The sensitivities of both acquisition methods should be similar for ~ 7.5²*10 = 560 s exposures. However, undispersed exposures will be superior for all reasonable acquisition exposure times up to ~ 100 s.

Specific predictions of the NIFS performance model for the H and K gratings are listed in Table 1 and Figure 1 and Figure 2 for 10 s exposures with 15 e read noise and 100 s exposures with 10 e read noise in both dispersed and undispersed light.

Table 1: NIFS Acquisition Performance Predictions

Magnitude	SNR (dispersed)		SNR (undispersed)
	10 s	100 s	10 s	100 s
H grating, Seeing=0.4, Strehl=0.4
15	52.9	…	444.7	…
16	18.2	…	191.2	…
17	12.0	107.3	60.8	…
18	3.8	40.6	30.1	108.2
19	2.3	20.2	12.5	32.4
20	…	8.3	4.6	11.3
21	…	3.0	…	7.1
22	…	…	…	3.5
K grating, Seeing=0.4, Strehl=0.6
14	124.4	…	…	…
15	53.5	…	323.9	…
16	26.0	…	140.7	…
17	9.8	104.4	67.4	…
18	3.5	42.3	22.0	79.2
19	…	17.9	9.0	29.4
20	…	8.5	3.9	14.7
21	…	…	…	5.7

Figure 1: H band acquisition performance for 10 s (solid) and 100 s (dashed), dispersed (diamonds) and undispersed (squares) exposures. The limiting magnitude at H is ~ 21.3 mag.

Figure 2: K band acquisition performance for 10 s (solid) and 100 s (dashed), dispersed (diamonds) and undispersed (squares) exposures. The limiting magnitude at K is ~ 20.3 mag.

5 Is Collapsing Dispersed Images Sufficient?

Collapsing dispersed images in the spectral direction to form an image of the sky is the simplest acquisition strategy to implement. However, the simulations in Figure 1 and Figure 2 suggest that it will fail by ~ 1 mag in acquiring objects at the sensitivity limit of the instrument in H band.

Limiting point sources will be detected with low signal-to-noise ratio in 100 s undispersed H band exposures.

We therefore conclude that it is necessary to include a grating mirror in the NIFS design. This demands that the grating be able to be reset to high accuracy in order to ensure that calibration frames remain valid.

Appendix A: List of Figures

Figure 1	snr.xls
Figure 2	snr.xls