Salter et al.

Direct Imaging Of Long Period Radial Velocity Targets With
NICI on Gemini South
Graeme S. Salter, Chris Tinney, Rob Wittenmyer (University of New South Wales)
Hugh Jones (University of Hertfordshire)
James Jenkins (Universidad de Chile)
Simon O’Toole (Australian Astronomical Observatory)
Graeme S. Salter
[email protected]
ABSTRACT We are finally entering an era where radial velocity and direct imaging parameter spaces are star=ng to overlap. Radial velocity measurements provide us with a minimum mass for an orbi=ng companion (the mass as a func=on of the inclina=on of the system). By following up these long period radial velocity detec=ons with direct imaging we can determine whether a trend seen is due to an orbi=ng planet at low inclina=on or an orbi=ng brown dwarf at high inclina=on. In the event of a non-­‐detec=on we are s=ll able TARGET SAMPLE AND AIMS The Anglo-­‐Australian Planet Search (AAPS) has been one of Most of the AAPS targets are 4-­‐6 Gy old G dwarfs, inac=ve the major Doppler search programs, opera=ng on the AAT and slowly rota=ng, providing good radial velocity stability. since 1998. It currently targets some 250 stars for regular We select all AAPS stars within 60pc, giving us 30 targets. Having such an old target sample means we are unlikely to observa=on and has detected over 40 exoplanets to date. Some 13% of our AAPS targets show evidence for long-­‐period be able to detect planetary mass bodies, however, a NICI companions at periods greater than the baseline of our data. non-­‐detec=on would provide an upper mass limit, pushing With their orbital periods only constrained to be longer than targets towards the planetary mass regime. the AAPS survey =me, we are unable to use the observed The detec=on of brown dwarf companions to these stars is veloci=es over that period to constrain the masses of these also of great interest. The AAPS targets are mostly G dwarfs, objects: they could have periods of ~20 years and orbital for which metallicity, gravity and age can be beYer separa=ons of ~6 au, and hence planetary masses, or they determined than any other class of star. Therefore, any could have periods of hundreds of years, separa=ons of brown dwarf companions discovered would break the age-­‐
>10au and hence brown-­‐dwarf or M-­‐dwarf masses. Direct mass degeneracy typically faced with un-­‐agable field brown imaging is the only means by which we can determine the dwarfs and spectra/photometry could be used as benchmarks in this field. nature of the long-­‐period candidates in our AAPS sample. 2:Par=al Annular 4:LOCI-­‐Like Fifng & 1:Calibrated Data Moving-­‐Minimum Fit Removal Frame 3:PAMF Removed (PAMF) Time = 1.4hours Time = 0.7hours Aher Aher AADI DI Before ADI <6.5(ΔHmag) @ 0.25” 5(ΔHmag) @ 0.72” 13(ΔHmag) @ 3.2” 4.5(ΔHmag) @ 0.5” ADI IMPLEMENTATION In order to reach the contrasts needed we have to perform Angular Differen=al Imaging (ADI), Marois ‘06. To do this the telescopes de-­‐rotator is switched off allowing the field to rotate on the detector but keeping the speckle noise sta=onary. In this way, the speckle noise can be fiYed and removed from the observed data. The panels in the lower leh show the 5 stages of our post-­‐
processing method. 1: Normal calibra=ons such as flat fielding, dark subtrac=on, bad pixel removal and de-­‐warping are performed. 2: NICI images show significant large scale structure . In order to fit this we find the minimum value within a par=al annulus (radial to the star) centered on each pixel. 3: The par=al annular moving-­‐minimum fit (PAMF) is then removed from the image, revealing the speckle noise. 4: We perform our own implementa=on of a Locally Op=mised Combina=on of Images (LOCI) analysis based largely on Lafreniere ‘07. For each image a PSF is made using the other images in the data set. Taking the image one small piece at a =me (AS) the PSF around each piece is scaled to match (AC). We reject up to two cores of pixels devia=ng significantly and re-­‐scale before subtrac=ng the found value. 5: Each image is de-­‐ Area of rotated and considera=on, AC Area of combined. subtrac=on, AS 5:All images De-­‐
rotated & Combined Time = 0 to put a limit on the maximum mass of the orbi=ng body. The Anglo-­‐Australian Planet Search (AAPS) is one of the longest baseline radial velocity planet searches in existence, amongst its targets are many that show long period trends in the data. We present here our direct imaging survey of these objects with our results to date. Imaging and ADI observa=ons have been made using NICI (Near Infrared Coronagraphic Imager) on Gemini South and analysed using an in house, LOCI-­‐like, post-­‐processing. 3.4(ΔHmag) @ 0.63” 15(ΔHmag) @ 6.4”(background object) 8.5(ΔHmag) @ 2.3” OBSERVATIONS AND PROGRESS Mean 5σ Mean 5σ The above plot shows radial profiles for the data, seen in the leh panels, before (blue) and aher (red) ADI post processing. The Y axis is in magnitudes fainter than the primary source. Actual pixel values are ploYed along with mean (dashed) and 5σ (solid) curves. All objects seen in these images were found to be background sources. 3 of our targets were not suitable for ADI observa=ons as their loca=on on the sky wouldn’t give enough field rota=on over a reasonable =me frame. For these targets we used regular dithered imaging. Pre-­‐imaging of some targets showed bright companions that have now been confirmed to be co-­‐moving so long ADI observa=ons were not required. To date, in our sample of 30, 11 targets have been found to have co-­‐
moving companions (a few examples of which can be seen above along with their corresponding radial velocity data). 12 targets have been shown to not have any companions observable with NICI. Through analysis of the achieved contrast (such as that seen leh) and inser=ng fake objects into the data, these observa=ons provide an upper mass limit to the system. A lower mass limit and a minimum orbital period is also provided by the radial velocity measurements. A further 7 targets have objects that have been detected but second epoch imaging is s=ll required to confirm that they are co-­‐moving. Results To Date: Target sample: 30 Targets with confirmed co-­‐moving companions detected: 11 Targets with no co-­‐moving companions detected: 12 Remaining targets with objects detected but co-­‐moving status s=ll unknown: 7