NICSPol: A Near Infrared polarimeter for the 1.2 m telescope at Mount Abu Infrared Observatory

TL;DR

NICSPol is a newly developed near-infrared imaging polarimeter for the 1.2 m telescope at Mount Abu, capable of measuring polarization with ~1% accuracy for sources brighter than magnitude 16 in JHKs bands, enabling diverse astrophysical studies.

Contribution

The paper introduces NICSPol, the first imaging near-infrared polarimeter in India, detailing its design, calibration, and performance on standard sources.

Findings

NICSPol can constrain polarization within ~1% for sources brighter than magnitude 16.

The instrument successfully observed polarized and unpolarized standards across J, H, and Ks bands.

NICSPol is suitable for studying various astrophysical objects like AGNs, pulsars, and supernovae.

Abstract

NICSPol is a near infrared imaging polarimeter developed for the Near Infrared Camera and Spectrograph (NICS), one of the back end instruments of the 1.2 m Cassegrain telescope at the Mount Abu Infrared Observatory (MIRO), India. The polarimeter consists of a rotating wire grid polarizer which is mounted between the telescope optics and NICS. The polarimetric observations are carried out by rotating the polarizer using a motorized mechanism to determine the Stokes parameters, which are then converted into the polarization fraction and polarization angle. Here we report the details of the instrument and the results of observations of IR polarimetric standards. A set of polarized and unpolarized standards were observed using NICSPol over J, H and Ks bands covering 0.8 to 2.5 micro m. The observations of polarized standards using NICSPol show that, NICSPol can constrain polarization within ~1% for sources brighter than ~16 magnitude in JHKs bands. NICSPol is a general purpose instrument which could be used to study variety of astrophysical sources such as AGNs, Pulsars, XRBs, Supernovae, star forming regions etc. With few NIR polarimeters available world-wide so far, NICSPol would be the first imaging NIR polarimeter in India.

Tables

Table 1. Table 1 : Specifications of NICSPol

Parameter	Specifications
Dimensions of WGP	WP25L-UB - 25.0 x 25.0 mm
Wavelength range	250 nm to 4 $\mu$m (useful range for NICS : 0.8 to 2.5 $\mu$m
Available NIR bands	Y (0.97 - 1.07 $\mu$m), J (1.17 - 1.33 $\mu$m),
	H (1.49 - 1.78 $\mu$m), K (2.03 - 2.37 $\mu$m), Ks(1.99 - 2.31 $\mu$m)
Detector	Teledyne H1RG detector with 1024 x 1024 arrays
Field of view	NICS - 8 x 8 sq arc min (imaging mode)
	NICSPol - 3.9 arc min dia (polarimetric imaging mode)
Pixel scale	0.5 arc sec per pixel
NICS optics and detector	Maintained at $\sim$ 77 K(Cooled by LN2)
WGP module	at ambient temperature
Limiting magnitudes	$\sim$ 16.8 in JHKs
	Typically with 30 - 40 sec exposure per frame
	and a net exposure of 30 min

Table 2. Table 2 : Log of all Observations using NICSPol

Source	Observed Nights	Filter	Positions	Frames	Orientations	Exposure/frame
HD 283809a	10	J	5	2 to 10	4	1 s
	7	H	5	2 to 5	4	1 s
	8	Ks	5	2	4	1 s
HD 204827a	5	J	5	3	4	801 ms
	4	H	5	3	4	801 ms
HD 29333a	5	J	4	2	4	1 s
	2	H	4	2	4	1 s
	1	Ks	4	2	4	1 s
HD 283725a	2	J	5	3	4	1 s
HD 283855a	2	J	5	3	4	1 s
HD 283701a	1	J	5	3	4	1 s
HD 56591b	1	J	3	3	11	2 s
HD 26212b	1	J	3	3	4	2 s
	1	H	3	3	4	2 s
	1	Ks	3	3	4	2 s
	1	K	3	3	4	2 s
	1	Y	3	3	4	2 s
NGC 2548b	1	J	5	5	4	20 s
M3b	1	J	3	5	4	30 s
HD 202573c	2	J	5	2	4	1 s
	2	H	5	2	4	1 s
	2	Ks	5	2	4	1 s
	2	K	5	3	4	1 s
HD 212311c	7	J	5	2	4	3 s
	5	H	5	2	4	3 s
	3	Ks	5	2	4	3 s
HD 103095c	3	J	4 to 8	5	4	801 ms
HD 65583c	1	J	7	5	4	801 ms
RA 07 - 08 hd	1	J	5	3	4	10 s

Table 3. Table 3 : Observed PF, PA for 100 % polarized standard stars.

Source	Band	Observed PF (%)	Obs PA (o)
HD 56591	J	100.75$\pm$0.52	-38.36$\pm$0.09
HD 26212	J	100.15$\pm$0.94	-38.45$\pm$0.24
	H	100.16$\pm$0.66	-38.37$\pm$0.18
	Ks	100.5$\pm$0.95	-38.21$\pm$0.24
	K	100.74$\pm$0.96	-38.20$\pm$0.20
	Y	100.58$\pm$1.22	-38.23$\pm$0.3

Table 4. Table 4 : Observed PF and PA for 100 % polarized light for the source NGC 2548.

Source	Obs PF (%)	Obs PA (o)
Star 1	100.45$\pm$0.62	15.24$\pm$0.16
Star 2	99.96$\pm$0.8	15.29$\pm$0.18
Star 3	100.66$\pm$1.67	15.43$\pm$0.39
Star 4	100.22$\pm$0.87	15.4$\pm$0.22
Star 5	99.66$\pm$0.36	15.49$\pm$0.10
Star 6	100.53$\pm$1.4	15.97$\pm$0.33
Star 7	100.44$\pm$0.8	15.53$\pm$0.21
Star 8	101.16$\pm$2.76	15.62$\pm$0.63
Star 9	99.57$\pm$0.27	15.44$\pm$0.07
Star 10	96.87$\pm$0.61	15.41$\pm$0.18
Star 11	96.12$\pm$1.37	14.21$\pm$0.35
Star 12	101.57$\pm$2.48	15.24$\pm$0.57
Star 13	100.9$\pm$1.15	15.49$\pm$0.29
Star 14	100.82$\pm$1.41	15.59$\pm$0.34
Star 15	99.04$\pm$3.63	14.88$\pm$0.83
Star 16	96.02$\pm$2.95	20.69$\pm$0.76
Star 17	98.00$\pm$2.85	18.57$\pm$0.73
Star 18	100.92$\pm$2.48	15.04$\pm$0.57

Table 5. Table 5 : Observed PF for Unpolarized standard stars.

Source	Band	Observed PF (%)	Ref
HD 202573	J	1.2$\pm$0.24	[33]
	K	0.21$\pm$0.23
HD 212311	J	0.65$\pm$0.68	[36]
	H	0.52$\pm$0.49
	Ks	1.37$\pm$0.81
HD 103095	J	0.39$\pm$0.23	[33]
HD 65583	J	0.07$\pm$0.52	[33]

Table 6. Table 6 : Standard PF, PA and Observed PF, PA for Polarized standard stars.

Source	Band	Std PF (%)	Obs PF (%)	Std PA (o)	Obs PA (o)	Ref
HD 283809	J	3.81$\pm$0.07	3.19$\pm$0.69	57$\pm$1	49.52$\pm$6.2	[32]
	H	2.59$\pm$0.07	2.12$\pm$0.33	58$\pm$1	56.29$\pm$4.4	[32]
	Ks	1.71$\pm$0.11	1.53$\pm$0.49	55$\pm$1	59.04$\pm$9.18	[32]
HD 204827	J	2.83$\pm$0.07	2.12$\pm$0.49	61.1$\pm$0.8	63.85$\pm$6.55	[37]
	H		1.71 $\pm$0.33		60.45$\pm$5.47
HD 29333	J	2.88$\pm$0.03	2.47$\pm$0.49	69$\pm$1	69.2$\pm$5.67	[32]
	H	1.81$\pm$0.04	1.81$\pm$0.49	68$\pm$1	70.57$\pm$7.73	[32]
	Ks	1.19$\pm$0.08	1.14$\pm$0.49	73$\pm$1	66.46$\pm$12.3	[32]
HD 283725	J	2.5$\pm$0.3	2.49$\pm$0.33	66$\pm$1	62.44$\pm$5.61	[32]
HD 283855	J	2.58$\pm$0.03	2.27$\pm$0.33	46$\pm$1	172.16$\pm$14.4	[32]
HD 283701	J	1.68$\pm$0.12	1.65$\pm$0.33	33$\pm$1.3	7.16$\pm$11.31	[34]

Equations

$${I=\frac{1}{2}(I_{0}+I_{45}+I_{90}+I_{135})}\ \$$

$${Q={I_{0}-I_{90}}}\ \$$

$${U={I_{45}-I_{135}}}\ \$$

$${I_{j}=\frac{1}{2}[I_{0}\pm Qcos2\theta_{j}\pm Usin2\theta_{j}]}\ \$$

$${PF=\sqrt{\frac{Q^{2}+U^{2}}{I}}}\ \$$

$${PA=\frac{1}{2}tan^{-1}\frac{U}{Q}}\ \$$

Full text

***Esakkiappan Aarthy, \linkableaarthy@prl.res.in , \linkableaarthy.e@iitgn.ac.in **

1 Introduction

Observational astronomy revolves around collection of electromagnetic radiation from various celestial sources. Spectroscopy, imaging and timing are the three most popular and commonly used techniques to analyze and interpret the photons received. There is a fourth comparatively less explored but important tool: astronomical polarimetry. Polarization is an inherent property of electromagnetic radiation which indicates the dominant orientation of the electric (& magnetic) field vector. Polarimetry gives two additional parameters: polarization angle and degree of polarization. Degree of polarization is the fraction at which the source is polarized. The orientation of the plane in which the electric vector oscillates, as it propagates, is the polarization angle. In general, astronomical sources are unpolarized unless there exist some asymmetry and/or anisotropy either in terms of matter or field, in the source or along the path of propagation of the photons. Hence measuring polarization helps in studying the magnetic field and geometry of various astronomical sources and their surrounding medium. Despite the benefits, polarimetry is relatively less popular compared to spectroscopy and imaging due to certain inherent issues. The prime reason is that most of the sources possess very low polarization (a few %) and hence longer exposure times are needed compared to spectroscopy.

The origin of polarization mostly differ over different parts of the electromagnetic spectrum. Polarization in UV, optical and IR are most commonly due to scattering by dust grains and magnetic field that orients the dust grains. Non thermal radiation from different sources are also polarized. In X-ray regime depending on the photon energy, geometry and magnetic field, polarization could be due to scattering (Rayleigh, inverse Compton), cyclotron, synchrotron [1] or due to more exotic physical processes like vacuum polarization and birefringence through extreme magnetic fields.[2],[3] In radio regime also non thermal radiation is synchrotron. The principle and methodology to acquire and analyze polarized data vary over the different regions of the spectrum. Assuming the initial radiation from the internal structure of the source to be unpolarized, measuring polarization is the only way to know about the point of interactions which polarizes it. The study of polarization induced by scattering is a powerful technique to understand the intervening matter and field distribution, when the original radiation is expected to be unpolarized (e.g. thermal radiation).

The Infrared regime of the EM spectrum helps in unveiling many embedded systems which could not be seen otherwise at optical wavelengths. The first measurement of IR polarization dates back to 1975 [4] and since then there have been several ground and space based instruments covering different regions from $\sim$0.8 $\mu$m in near IR to few tens of $\mu$m in the far IR studying wide range of astronomical sources. Few well known near Infrared polarimeters are: SIRPOL at the 1.4 m IRSF [5], Mimir at the 1.8 m Perkins telescope [6], NICMOS at the 2.4 m HST [7], MMTPOL at the 6.5 m MMT Observatory [8], SOFI at the 3.58 m NTT telescope, La Silla Observatory [9], POLICAN at the 2.1 m telescope of the Guillermo Haro Astrophysical Observatory (OAGH) located in Cananea, Sonora, Mexico [10], TRISPEC as a visitor instrument at several facilities including UKIRT, UH, OAO, Subaru telescope [11], SPHERE at the 8.2 m VLT [12], HONIR at the 1.5-m Kanata telescope [13] etc.

We have recently added a polarimeter NICSPol to the existing Near Infrared Camera and Spectrograph (NICS) which is one of the back end instruments for the 1.2 m Cassegrain f/13 telescope located at the Mount Abut Infrared Observatory of Physical Research Laboratory [14]. This makes NICSPol the first imaging IR polarimeter in India. NICS is capable of doing both photometry and spectroscopy covering a wide range of 0.8 to 2.5 $\mu$m (Y, J, H, Ks) and has been serving as a work horse for the past several years to study diverse objects like AGN, galaxies, supernovae, novae and compact objects etc. Kolokolova et al, 2015 [15] and other excellent review articles [16], [17], cover polarimetry and the scientific aspects over a wide electromagnetic domain. Polarization of light in infrared could be majorly due to scattering (e.g. polarization in comets, planets, asteroids etc are due to scattering by dust grains), and/or non thermal radiation (e.g. synchrotron radiation from AGN, XRBs). The scientific prospects of IR polarization in case of both galactic and extra galactic sources have been extensively reported in the literature [18], [19], [20], [21], [22]. Having optical polarization measurement with the 50 cm telescope of PRL [23], and the hard X-ray polarization capability of AstroSat CZTI [24], NICSPol would provide a fantastic opportunity for simultaneous polarimetry over a wide range of EM spectrum.

Here we report the NICSPol instrument design and its calibration results for IR polarimetric standards. Different polarization measurement techniques are briefed in Section 2 and NICSPol instrument specifications are given in Section 3. Sections 4 and 5 describe the observations and the data reduction. The results are mentioned in Section 6 followed by the summary in Section 7.

2 NIR Polarimetry

2.1 Instrumental Techniques

The means of acquiring the polarimetric information varies depending on the energy regime of the incoming photons. There are several methods to measure polarized radiation from astronomical sources. The efficiency of the measurement method is one of the deciding factors in preferring a particular scheme. Using a Wollaston prism along with a rotating half wave plate modulator, gives both the orthogonal polarization states in a single exposure, hence also building up on the total efficiency of the system [25]. WeDoWo (Wedged Double Wollaston) is another similar technique with the difference that two Wollastons are combined together, giving images at four angles (0o, 45o, 90o, 135o) simultaneously in a single exposure [26]. This technique, hence, helps in simultaneous observation and prevents variations in the sky background during the exposure, from affecting the results. The third one and the one used in NICSPol is a Wire-grid Polarizer (WGP). It works on the principle of dichroism, and functions as an absorptive polarizer. The selective direction in the case of WGP is perpendicular to the wire-grids, resulting in linear polarization. The wavelength range being covered by a WGP is dependent on the spacing and widths of the wires in the WGP [27]. The extinction ratio of polarizers is a measure of the ability to attenuate light in direction perpendicular to transmission axis of polarizer[28].This is an important parameter which decides the efficiency of the polarizers. This ratio for WGP in infrared has a value of 1000:1. Compared to other polarization optics components, WGP can be made in larger apertures. Thus, WGPs are effective polarizers with large field of view (FOV), are fairly compact and have good stability [27]. Hence, they are used extensively in polarization measurement. Our choice for WGP method of polarization measurement in NICSPol is also motivated by the limited space available in the light path between the telescope Cassegrain mounting flange and the instrument entrance aperture.

2.2 Analysis Technique

The polarization of a source is defined by two parameters, the polarization fraction (PF) or the degree of polarization and the polarization angle (PA). There are different methods to acquire these two information among which the most preferentially used method in optical or IR is the Stokes method. When a polarized beam passes through an analyzer which is rotating at discrete steps, the output beam gets modulated and follows a Cos2$\theta$ distribution. A mathematical fit (equation 4) to the observed intensities provides the PF and PA. But getting intensities of a source at multiple steps over 0o to 180o is quite time consuming and in the case of the infrared domain the assumption that the sky has not changed may not hold. The advantage of Stokes method is that, it requires intensities only at 3 (or 4) different orientations and effectively provides the PA and PF using any 3 (or all 4) of the I0, I45, I90, I135 (intensities obtained by rotating the WGP in steps of 45o) measurements. Given by GG Stokes in 1852 [29] and introduced in astronomy by S Chandrasekhar in 1947, Stokes parameters I, Q, U, V describe the polarization state of a system in terms of intensities [30]. I (equation 1) represents the total intensity, Q (equation 2) and U (equation 3) represent linear polarization and V represents circular polarization. The Stokes parameters are calculated from the observations of flux measured at different orientations of the WGP. The following equations give the relation between intensity and Stokes parameters [31].

[TABLE]

[TABLE]

[TABLE]

Alternatively, Stokes parameters could be obtained by by fitting equation 4,

[TABLE]

The PF and PA are obtained from the Stokes parameters using equations 5 and 6.

[TABLE]

[TABLE]

3 NICSPol Instrument description

NICSPol consists of a 25.0 x 25.0 mm WGP (WGP) commercially available from Thorlabs, which covers a wavelength range of 250 nm to 4 $\mu$m. The layout of the NICSPol instrument is shown in Figure 1. Figure 2 shows the instrument mounted on the Cassegrain plate of the MIRO telescope. The WGP mounting details are also shown in this figure. NICS is capable of doing both photometry and spectroscopy by the use of mirror and grating. The specifications of NICS are given in Table 1. NICS consists of a filter wheel with Y, J, H, K, Ks filters. The detector in NICS is Teledyne H1RG detector with 1024 x 1024 arrays which is cooled by using liquid Nitrogen and covers a FOV of 8 x 8 sq arc min [14]. At the top of NICS, the part which is attached to the instrument ring of the telescope is a 20 cm x 20 cm square box which contains a beam splitter. This essentially sends a fraction of the light from the source to a guiding ccd which helps in telescope tracking. The use of a warm WGP is necessitated by the requirement that the polarization module could be inserted in (or removed from) the light path without fully dismounting the entire instrument. The only drawback in this scheme (warm WGP) is the increased thermal background in the K band which restricts us to use only the Ks filter in the longer wavelength side.

An Optec Pyxis LE Camera Field Rotator was found suitable to mount the polarizer. The field rotator has a barrel which is rotated by a stepper motor with a step size of 1o rotation. A delrin module was made to hold the polarizer firmly and the module was mounted in the barrel of the field rotator. A side of the beam splitter holder box was replaced by the module as in Figure 2 at a proper position so that the polarizer is exactly along the line through which the light from the secondary reaches the NICS optics. The polarimetry and imaging/spectroscopic modes are easily interchangeable. Since the size of the WGP is such that it blocks some of the incoming beam, there is strong vignetting (see Figure 3) beyond half-field of 1.95 arcmin. This limits the useful FOV to 3.9 arc min (dia).

The faintest stars seen using NICS in J, H, Ks bands, using our 1.2 m telescope reach magnitudes of $\sim$17 with individual exposures of 30 to 40 sec per frame and a net exposure of 30 minutes. However, the introduction of the WGP will result in brighter detection limits, due to the $\sim$85% transmission of the WGP in NIR [28] making the NICSPol limiting magnitude $\sim$16.8. However in polarimetric mode we expect that these magnitudes might be achieved only under very stable skies.

4 Observations

This section describes the list of observations made to verify the instrument performance using lab and sky polarimetric standards. The instrument had to be checked with polarized and unpolarized standards for efficiency and instrumental polarization respectively. NICSPol was calibrated using both polarized and unpolarized stellar standards. A second WGP of same dimensions was also used to achieve 100 % polarized light. This WGP was mounted stationary in NICSPol in a way that light from the source would first pass through it, get polarized and pass through the rotating WGP resulting in a 100 % polarized light. For calibration with stellar sources both isolated stars and crowded fields were observed.

Table 2 shows the complete log of all the observations made for the calibration of NICSPol. A major issue faced was the small number of stellar polarimetric standards in NIR. The few available standards are polarized to a maximum of 3 - 4% in J band. 6 UKIRT polarimetric standards : HD 283809, HD 204827, HD 29333, HD 283725, HD 283855 and HD 283701 and four unpolarized standards : HD 202573, HD 212311, HD 103095 and HD 65583 (to check for instrumental polarization) [32],[33] were selected based on their availability at MIRO winter sky. The polarized standards were chosen such that they possess a degree of polarization of about 2 - 4 % and these were followed up over multiple nights from September to November 2017 and December 2018 to January 2019. A critical part of NIR observations is the removal of sky contribution, which is bright in infrared. For a particular source, for each filter the polarimetric images were taken by dithering the source in minimum 5 positions and the sky frame was made from these multiple dithered frames of the source itself. In case of extended sources sky frames were taken separately by taking the source out of the field of detector for the same exposure as the source, assuming the sky to be constant over this dithering period. The source images were taken for 3 filters x 5 dithering positions (with 3 frames per position) x 4 orientations of the WGP. The exposure time varied from 801 ms in case of bright point sources (minimum achievable by NICS) to 20 s each for faint sources. Also to check how closely the results from 4 orientations would match with the result from data with multiple points fitted using equation 4, HD 56591 was observed with 2 WGPs (i.e. polarizing the light from the source), for 0 o to 150 o with 15 o angular step size.

5 Data Reduction and Analysis

As explained in Section 4 the Stokes method is followed to analyze the polarimetric data. The analysis was carried out using standard IRAF (Image Reduction and Analysis Facility) procedures along with few IDL scripts to ease the analysis. The initial steps to analyze the polarized images in IR are the same as standard IR photometry. For a given source and a given band, the images had been taken by rotating the polarizer in four orientations, 0o, 45o, 90o, 135o and multiple frames are taken for each dithering position. And as mentioned earlier the IR sky is bright hence sky subtraction is a critical step in the analysis. Sky frames were constructed by median combining source frames taken at different dithering positions. This sky image obtained was subtracted from all the raw source images to get the sky subtracted frames. To improve the S/N ratio multiple frames of all the positions should be added up. Since the position of the source would be different in each frame, it was shifted to a common point with respect to a particular star and the shifted images were combined to give a final image for each orientation. The next step is to get the intensities of the source at the 4 orientations. Similar to photometry the magnitude of the source should be extracted and then converted into intensity. For this the IRAF phot procedure was used to do aperture photometry, where a range of apertures were defined for the source and using the curve of growth plot a suitable aperture size was considered for further analysis. The above procedures were repeated for all 4 orientations of observations hence giving 4 intensities I0, I45, I90, I135. The Stokes parameters were calculated from these intensities and the degree of polarization and the polarization angle are obtained from the Stokes parameters. The additional advantage of following Stokes method is that since normalization is done with Intensity (I) any systematic error in the parameters are cancelled out. In the case of extended sources a standard IRAF procedure, linpol, would directly give pixel by pixel information of PF, PA, I, Q, U.

6 Results

6.1 100 % Polarized Light

By mounting a second stationary WGP the incoming light from a source is polarized, using this the polarimetric efficiency of NICSPol was tested by observing HD 56591, HD 26212, NGC 2548 and M3. The obtained PF and PA are tabulated in Tables 3 and 4. The modulation curves for HD 56591 and HD 26212 are plotted in Figure 4. The reason behind the PF crossing 100% within the error bar is that, the intensity of the source at 135o and above is very less which increases the uncertainty in the photometric measurement. The fitted PA of both the sources are found to be consistent over all the bands and the error in PA is found to be only $\sim$ 0.1 - 0.2o. Figure 5 shows the NGC 2548 field marked with polarization vectors of all the sources whose details are tabulated in Table 4. It could be seen that the results are consistent over the entire FOV of the NICSPol module over the detector.

6.2 Unpolarized Stars

Four unpolarized standards - HD 202573, HD 212311, HD 103095 and HD 65583 were observed using NICSPol and the obtained PF are tabulated in Table 5. HD 103095 and HD 65583 was also observed by dithering the source at multiple positions (8 positions). Figure 6 shows the PF and PA obtained over various positions in the FOV for both HD 103095 and HD 65583. The PFs at all the positions were lesser than 1%. and also the PAs over these positions were found to be random. Figure 7 shows the modulation of unpolarized standard HD 103095. Figure 8 shows the PF of one unpolarized standard HD 212311 observed over several nights which is found to be around 1%. From unpolarized standards it is seen that the uncertainty is around $\sim$1%.

6.3 Polarized Standard Stars

IR polarized standards are less in number and the degree of polarization is generally quite low compared to optical wavelengths. Six polarized standards - HD 283809, HD 204827, HD 29333, HD 283725, HD 283855 and HD 283701, were observed and few of these were followed up during several observing nights. The obtained PF and PA along with standard reported values are tabulated in Table 6. The references of reported PF, PA of chosen standards are given in Ref column. The instrumental polarization angle obtained by fitting Equation 4 has to be converted into the standard polarization angle with reference to the local North in the equatorial coordinate system. Over a given observation night the difference between the instrumental PA and Standard PA is expected to be same for different sources and by implementing this correction factor Observed PA is obtained. Figure 9 shows the difference between the instrument and standard PA for all the sources over all observation nights. It could be seen that the observed values match with the standard values well within the error bar for sources HD 283809, HD 204827, HD 29333 and HD 283725. Figure 10 shows the PF and PA values obtained for HD 283809 with the standard values quoted inside for J, H and Ks bands and modulation curve of one observation is given in Figure 7. The observed PF and PA for HD 283809 and HD 29333 were found to be consistent when focused over different positions of the detector FOV. For sources HD 283855 and HD 283701 although the Observed PF match the Standard PF, the Observed PAs are different from the reported values. It should be noted that at least for HD 283701, different PA values of 42o$\pm$5o (Whittet et al. 1992 [32]) to 33o$\pm$1.3o (Whittet et al. 2001 [34]) are reported in literature. This indicates the possibility that the PA change could be real, which can only be confirmed by further observations that we plan to continue. Figure 11 shows the obtained PF and PA plotted with respect to the reported values of PF PA. MIMIR instrument reported the PF and PA of standard stars with small error bars [6]. The reason for our relatively large errors in comparison with MIMIR is, in MIMIR’s case the observations are taken for 32 half wave plate position angles, i.e. for every 11.25o. Also MIMIR is at 1.8 m Perkins telescope hence has large number of photons available therefore less photon noise. These result in higher precision in the measurements. For e.g. in one of our 100% polarization observations (6.1), one set of observations were made for every 15o position angle covering 0o - 165o (Figure 4). From these 11 position angles of the WGP PF = 100.75%$\pm$0.52%, PA = -38.36o$\pm$ 0.09o are obtained. Whereas when only 4 position angles from these with a stepsize of 45o was used the values obtained are PF = 100.8%$\pm$0.9%, PA = -38.49o$\pm$0.2o. The increase in errors due to the smaller number of position angles is evident in the case of 100% polarized light and this would affect the sources with low polarization. Hence in the case of NICSPol it is preferable to go for more position angles while observing sources with low PF, while 4 positions angles would be sufficient to study highly polarized sources.

6.4 Polarization Measurement of a Photometric Standard Field

Landolt, A. U. in 1992 [35] listed photometric standards around the celestial equator whose V band magnitudes range between $\sim$ 11 - 16. RU 149 a blue star, and other surrounding stars (RU 149 A - G) in the field which are in RA 07 - 08 h was observed using NICSPol. Figure 12 shows the field observed and by comparing with the standard 2mass image, the magnitudes of the stars in J band were obtained which are marked in the figure. To estimate the limiting magnitude which could be achieved using NICSPol, polarization analysis of all the stars in the field was carried out. The exposure time per frame was 10 s (5 positions, 3 frames each, 4 angles) which resulted in a net exposure of 10 min for the complete observation. It could be seen that with this net exposure it is possible to measure polarization for stars as faint as $\sim$15 mag in J. Hence with a higher exposure it is possible to reach up to $\sim$16.5 (fainter stars seen in the field) using NICSPol. This matches very well with the number quoted earlier in Section 3 ($\sim$16.8). Since the observed field is closer to the Galactic plane the stars in the field can not be concluded as unpolarized.

7 Summary

We have presented here the design and test results of NICSPol, an NIR polarimeter add-on for NICS instrument covering wavelength range of 0.8 - 2.5 $\mu$m for the 1.2 m telescope at MIRO. NICSPol is the first imaging polarimeter in India and covers a FOV of 3.9 arc min dia and has a pixel scale of 0.5 arc sec per pixel. The polarization analysis is performed using IRAF and IDL following the standard Stokes methodology to obtain the degree of polarization and the polarization angle. Observations of NIR polarimetric standards as well as unpolarized stars using NICSPol show that the polarization degree and angle can be accurately determined within 1% and few degrees, respectively. This shows that NICSPol is suitable to carry out NIR polarization of sources with magnitudes brighter than $\sim$16 over J, H, Ks bands and polarization greater than $\sim$1%. As discussed in section 6.4, for highly polarized sources, observations with large step angles would be sufficient, however for sources with low polarization, we may use more steps to fill the modulation curve and achieve higher accuracy. NICSPol is a general purpose instrument at MIRO and can be used for observations of both point and extended sources. In future, we plan to use NICSPol simultaneously with the optical polarimeter available at the 50 cm telescope at Mt Abu to study wide variety of galactic and extra galactic sources and hence many more scientific results are expected in near future.

Acknowledgements.

Contribution of PRL workshop is acknowledged for their support in making required parts and the overall support involved in mounting the WGP between the telescope and the NICS instrument. We thank the observatory staff at MIRO for their technical contributions and for the support of the observations. We thank Mohan Lal for his technical help during the prototype machining of WGP support for initial tests of NICSPol concept. We also thank Aravind K and Prashanth Kumar Kasarla for their help in making the delrin mount for the polarizer. The authors thank the telescope operator trainees Prashant Chauhan and CM Mukesh for their timely help during the observations.