Arpita Roy

Dept of Astronomy & Astrophysics, Penn State University

[Moved to the caltech as a Robert A. Millikan Prize Postdoctoral Fellow]

The Lunar Farside Highlands Problem

The Moon exhibits a dramatic dichotomy between hemispheres. The familiar nearside of the Moon, constantly visible from the Earth, is largely low-lying, flat, and buried in volcanic maria. The farside consists mostly of mountainous highlands and bears witness to extensive cratering. The two lunar faces thus have different topographies, with the highest elevations found in the farside highlands. In fact, the farside crust is thicker than the nearside crust. Why? We think the solution involves Earthshine, and "metal snow" on an infant Moon.


NEID (“to see” in the language of the Tohono O’odham) is an instrument designed in response to the NASA-NSF (NN-EXPLORE) call for an extreme precision spectrograph for the astronomical community. NEID is being built by a multi-institutional and interdisciplinary team led by Penn State, and will be delivered to the WIYN telescope in 2019. As part of the NEID team, I have an important role in instrument design, and am co-leading the software development effort. This instrument will really push the boundary on current precision-limiting factors, and will define the gold standard in the field until we have access to next-generation instruments on extremely large telescopes.

The Habitable Zone Planet Finder

The Habitable Zone Planet Finder (HPF) is a high-resolution fiber-fed near-infrared spectrograph that will be commissioned at the 10m Hobby-Eberly Telescope (HET) in Texas in early 2017. Instrument subsystems are currently being assembled and tested at Penn State. I am co-leading the software development for HPF, which includes a search for more reliable stellar activity indicators in the NIR. I am also heavily involved in the design and fabrication of the optical fiber feed that will transport light from the telescope focal plane to the instrument. R&D areas include stabilizing illumination via a ball lens scrambler (patent pending), modal noise mitigation, and minimizing focal ratio degradation while maintaining sufficient throughput.


PARAS is a fiber-fed stabilized high-resolution cross-dispersed echelle spectrograph, located on the 1.2m telescope in Mt. Abu India. Designed for exoplanet detection, PARAS is currently achieving radial velocity (RV) precisions approaching ~1m/s over several months. Workhorse spectrographs like PARAS are essential for concerted exoplanet follow-up and provide critical field-experience for the development of future cutting edge instruments. I wrote and maintain the reduction and RV analysis pipeline for PARAS, and am using it to explore a variety of science and instrumentation goals.

Stellar Activity

In an era of prolific planet detection, high-resolution spectra provides the essential ability to monitor spectral line shapes and distinguish the true bulk motion of a star due to a planetary companion from false positives, including activity and contamination. I use metrics of the cross-correlation function to measure stellar activity and simulate possible contamination scenarios. This technique provided the breakthrough in the insidious case of MARVELS-1b: the face-on double-line binary that was masquerading as a resonant planetary system. It has also proved useful in several cases where stellar activity was mimicking the signature of promising habitable-zone planet candidates (Gliese 581, Kapteyn’s star).

Reflected Light from Exoplanets

Groundbreaking work in the last decade has led to the direct detection of atmospheric signatures of exoplanets. However, the field of exoplanet characterization is still in its infancy, with only ~50 detected exoplanet atmospheres, severely limiting theoretical models of comparative exoplanetology. I plan to use stabilized high-resolution spectroscopy to detect the reflected light from transiting and non-transiting exoplanets in the optical. Reflected light provides manifold scientific gains and can yield the true mass and rotation rate of a planet, the systemic inclination, and perhaps most interestingly, the planetary atmospheric albedo. Given the complex and often disequilibrium nature of planetary atmospheres, this work will provide critical observational constraints on atmospheric models, in anticipation of the next generation of instruments for exoplanet characterization (e.g. JWST, ESPRESSO, HIRES, G-CLEF).

Combining RV + Astrometry

Radial velocity (RV) measurements do not constrain the inclination of a system and can only provide a lower limit to the companion mass. However, RV can be combined with complementary observations like transit photometry, or the astrometric motion of the host star, to resolve this ambiguity and recover the 3D configuration of a system. I examine stars in the Hipparcos catalog, and reconstruct existing astrometric data constrained by reliable spectroscopic parameters. While Hipparcos is limited to the detection of approximately brown-dwarf mass objects, these systems prepare us for GAIA, which will enable the detection of planetary companions.


I am a graduate student at Penn State working on my PhD in Astrophysics. My research focuses on improving spectroscopic techniques for the detection of exoplanets, in order to advance us into the realm of extreme precision spectroscopy necessary for the discovery of Earth analogs. I am part of three instrument and science teams: HPF, NEID, and PARAS. I lead/co-lead the software development for all three instruments, and am deeply involved in overall instrument design. I use these high-quality spectra to improve radial velocity measurements, study stellar activity and push on direct detection methods from the ground.

Curriculum Vitae {Updated September 2016}
Contact Me
Parallax midground clouds
Parallax background clouds
cloud cloud cloud cloud cloud
Landscape with trees and cows cloud cloud cloud