DIRAC Postdoctoral Positions

The DIRAC Institute in the Department of Astronomy at the University of Washington is seeking applicants with a strong research record in the development of statistical techniques or algorithms for analyzing large astrophysical data sets for two postdoctoral positions.

AstroML: The first position is to help in the development of the second edition of astroML (http://astroml.org) a popular Python-based machine learning package for astrophysics. New components we are incorporating within astroML include methodologies from deep learning and hierarchical bayesian statistics. Special emphasis will be placed on building a broader community and making astroML a sustainable open-source project. The successful candidate will lead these activities, including the application of the new codes to dataset available to UW researchers.

Time Series Data: The second position is to develop new approaches for analyzing astronomical time series data using modern computational frameworks. The goal of this framework will be to enable science with the ZTF and LSST data sets. Promising applicants should possess an interest in time domain science and experience or interest in the use of databases and large scale compute platforms such as Spark, Dask, or similar. Good Python skills, and experience with machine learning libraries, image processing of astronomical images, or astronomical databases are desirable.

The DIRAC Institute is a newly formed center for data intensive astrophysics at the University of Washington. The Institute consists of six faculty and senior fellows, and over 20 postdoctoral researchers and research scientists. It has active research programs in Cosmology, Solar System science, Milky-Way structure, the Variable and Transient universe, andAstronomical Software.

The University of Washington is a partner in the Zwicky Transient Facility (ZTF) project, a new time-domain survey which will begin operations in early 2018. The UW is a founding partner of the LSST project, and leads the construction of its time domain and solar system processing pipelines. Other research activities at UW/DIRAC include topics in extragalactic science, as well as the understanding the structure, formation, and evolution of the Milky Way using large surveys (SDSS, WISE, PanSTARRS PS1, and others).

A Ph.D. degree in astronomy, physics, computer science, or a related subject is required. The initial appointment is for two years, renewable up to three years, and offers competitive salary and benefits. The appointments are available immediately and are expected to start no later than September 2018.

Applicants should submit a curriculum vitae, description of research interests (with links to Github if relevant) and arrange for three letters of reference to be submitted to Nikolina Horvat at horvat@uw.edu with subject line “DIRAC postdoc application (your name)”. Applications will be accepted until the positions are filled, to assure full consideration, please send your application by Dec 31st 2017

For detailed information about the benefits available through the University of Washington, including dental, medical and disability insurance, retirement, and childcare centers, see the University of Washington benefits page: https://www.washington.edu/admin/hr/benefits/.

The DIRAC Institute is a community of people with diverse interests and areas of expertise, engaged in the understanding of our universe through the analysis of large and complex data sets. We are an open, ethical, highly engaged and collaborative community based on trust, transparency and mutual respect. We believe in providing a welcoming and inclusive environment, in the importance of quality of life, in embracing diversity, in making a difference and having fun.

2019-10-14 Seminar: François Lanusse

When: October, 14, 2019 @ 12:30pm Where: PAB, 6th Floor, eScience Studio, Seminar Rm.

Deep Generative Models of Galaxy Morphology and Application to Deblending

The upcoming generation of wide-field optical surveys which includes LSST will aim to shed some much needed light on the physical nature of dark energy and dark matter by mapping the Universe in great detail and on an unprecedented scale. However, with the increase in data quality also comes a significant increase in data complexity, bringing new and outstanding challenges at all levels of the scientific analysis. In this talk, I will illustrate how deep generative models, combined with physical modeling, can be used to address some of these challenges at the image processing level, specifically by providing data-driven priors of galaxy morphology.

I will first describe how to build such generative models from corrupted and heterogeneous data, i.e. when the training set contains varying observing  conditions (in terms of noise, seeing, or even instruments). Once trained, sampling from these models produces realistic galaxy light profiles, which can then be used in survey emulation, for the purpose of validating and/or calibrating data reduction pipelines.

Even more interestingly, these models can be seen as priors on galaxy morphologies and used as such as part of standard Bayesian inference techniques to solve astronomical inverse problems ranging from deconvolution to deblending galaxy images. I will present how combining these deep morphology priors with a physical forward model of observed blended scenes allows us to address the deblending problem in a physically motivated and interpretable way.  

About François Lanusse

I am currently a postdoctoral follow at the Berkeley Center for Cosmological Physics (BCCP) and with the Foundation of Data Analysis (FODA) institute at UC Berkeley, where I conduct my research at the intersection between cosmology and machine learning.

Previously, I was a postdoctoral researcher in the McWilliams Center for Cosmology at Carnegie Mellon University, where I was working with Prof. Rachel Mandelbaum on weak gravitational lensing measurements and systematics, and also interacted with both Statistics and Machine Learning departments here at CMU.

I did my PhD in the CosmoStat laboratory of CEA Saclay near Paris, France, under the supervision of Jean-Luc Starck. My PhD work focused on the application of sparse regularization techniques to solve ill-posed inverse problems in a cosmological context.

Before that, I received a Master’s degree in fundamental physics from Paris-Sud University as well as a Master’s degree in fundamental and applied mathematics from Paul Verlaine University in Metz (France). I am also a Supélec engineer, from one of France’s top grandes écoles, where I specialized in Robotics and Interactive Systems.

You can find my academic CV here.

DIRAC Seminar Series

DIRAC Seminars are appropriate for students and researchers at all levels who are interested in astronomy and/or data science, and everyone is welcome to attend. They are not typically streamed or recorded.

Where: PAB, 6th Floor, eScience Studio, Seminar Rm. Time: 12:30pm | Seminar dates are listed below

Talk titles, abstracts, opportunities to meet with speakers, and information about lunch will be publicized on the UW Astronomy “astro-all” email list and the DIRAC Slack the week before each talk. Most speakers will visit DIRAC for a day or two and be available for meetings following their talk. For more information, you can contact DIRAC Visitor Committee Chair Stephen Portillo at sportill@uw.edu.


OCT 14 | Francois Lanusse | UC Berkeley | deep learning and cosmology

NOV 12 | Ashley Villar | Harvard | machine learning and transient light curves

DEC 19 | Adam Smercina | U Michigan | galaxy formation and detecting faint diffuse structure


JAN 13 | Michael Medford | UC Berkeley | IMBH and Planet Nine searches with ZTF

FEB 10 | Stephanie Juneau | NOAO | galaxy and SMBH evolution using large surveys

MAR 9 | Angus Wright | U Bonn | redshift calibration and galaxy clustering estimation

APR 13 | Bruce Conway | U Illinois Urbana-Champaign | astrodynamics and heuristic optimation

MAY 11 | Abigail Stevens | Michigan State U / high energy observations and time series

JUN 8 | Federica Bianco | U Delaware | LSST science collaborations chair and transients and variable stars science committee chair

2019-09-24 Seminar: Kaushal Sharma

When: September 24, 2019 @ 12:00pm

Where: PAB, 3rd floor, B305

Classification in Astronomy using ML.

Dr. Sharma is a postdoc at the Inter University Centre for Astronomy and Astrophysics in Pune, India. His Tuesday noon talk will be on “Classification in Astronomy using ML.” In his research, he applies machine learning and deep learning to the classification of stellar spectra and LMXB X-ray spectra.

Abstract: The huge data volume generated from the ongoing and upcoming astronomical survey programs has pushed the astronomers to shift from the traditional methods to more sophisticated and scalable approaches for data reduction and analysis without compromising with the precision and accuracy of the results. To address these concerns, Machine Learning (ML) algorithms have been considered in Astronomy, like other domains, and have been employed successfully for a wide range of astronomical problems in the last couple of decades. Classifying astronomical objects is one such field and in this talk, we will consider:
(1) stellar spectral classification using machine learning (ML) and deep learning (DL) techniques like ANN, RF, and CNN based on their spectra in the optical region of EM spectrum. We show that using CNNs, we are able to lower the RMS error up to 1.23 spectral sub-classes and apply the final model on stellar spectra from the SDSS.
(2) classification of X-ray spectra of Low Mass X-ray Binary (LMXB) systems which consist of a main-sequence star and a compact object that could be a Black Hole (BH) or a Neutron Star (NS). Using random forest (RF) algorithm, we are able to identify the type of compact object in the LMXB system (BH or NS) by feeding the X-ray energy spectrum to the model with an average accuracy of 87+/-13%.

Letter from the Director

A total solar eclipse occurred on July 2, 2019 over Cerro Pachón in Chile.

Welcome to our second DiRAC Institute newsletter!

The image above shows the solar eclipse that passed over the LSST site on July the 2nd. This remarkable event culminates a number of important milestones for the LSST over the last year. The 8.4m M1M3 mirror (the main mirror for the LSST) was delivered by ship to Coquimbo, a port in Northern Chile, and then driven to the LSST’s summit where it awaits the arrival of the telescope that will hold the mirror. The dome and building for the LSST, as you can see in the image, is nearing completion with the skin of the dome being carefully put into place. 

Some of the most interesting events we might find with the LSST, and are already finding with ZTF and the University of Washington telescopes at Apache Point Observatory are the visible counterparts to gravitational wave detections. After many decades of research and development the first gravitational wave was detected in 2015 and the first optical counterpart to a gravitational wave in 2017. Now with the new capabilities of LIGO (Laser Interferometer Gravitational-Wave Observatory ) gravitational waves from distant astronomical sources are being detected multiple times a month. Searching for the optical counterpart is tricky as it is hard to localize the source of the gravitational waves on the sky and only some of the sources will be visible in the electromagnetic spectrum and even then for only a short amount of time. DiRAC scientists Zach Golkhou, Mellisa Graham, and Eric Bellm are part of the Growth Project to followup the gravitational wave detections and are bringing the computational skills associated with DiRAC to bare on this problem. Zach talks about the excitement and challenges of this work in one of our featured highlights below. 
Closer to home, in as much as anything in astronomy can be considered close to home, June the 30th was Asteroid Day and we celebrated the event with a video of some of the work we are doing at DiRAC to explore our Solar System by studying the properties of the populations of asteroids and comets. We hope you will enjoy seeing some of our undergraduate and graduate students, postdocs, research scientists, and faculty and the work they are doing. Please consider supporting DiRAC and our mission as we look forward to better understanding the universe and how it came into being.

DiRAC Scientists at the Asteroid Day 2019

During the last week of June, Asteroid Day celebrated their Fifth Anniversary in 2019, with events in 192 countries, and once again broadcast their six-hour Asteroid Day LIVE TV show from Luxembourg, hosting innovators, astronauts, planetary scientists, celebrities and asteroid experts.

Scientists from DiRAC Institute and LSST, Prof. Mario Jurić and Dr. Lynne Jones, were among the panel guests. The program featured two short films about team’s work and current research.

DiRAC & LSST videos were broadcasted under names: 2017 | The LSST Making Census of the Solar System, 2019 | New Era of Cosmic Discovery

“Asteroids” Trailer from Nikolina Horvat on Vimeo.

New Era of Cosmic Discovery from Nikolina Horvat on Vimeo.

Meet DiRAC’s Research Team: Dr. Sarah Greenstreet

Sarah Greenstreet’s research interests involve orbital dynamics of small bodies in the Solar System. This includes studying the long-term evolution of asteroids in the inner and outer Solar System as their orbits change over time and determining their potential to impact the planets. Dr. Greenstreet’s current main focus is working with researchers at the Asteroid Institute as well as Associate Professor and Senior Data Science Fellow Dr. Mario Juric (DIRAC, Department of Astronomy, eScience) to study the threat to Earth due to asteroid impacts and by how much we would need to “nudge” an asteroid out of Earth’s way to avoid an impact.

Dr. Greenstreet is an Asteroid Institute senior researcher and a visiting scientist at the DIRAC center. She joined DIRAC in January 2018 after completing a two-year postdoctoral fellowship at Las Cumbres Observatory in Santa Barbara, CA. She received her Masters in Astronomy in 2011 and her PhD in Astronomy in 2015 from the University of British Columbia in Vancouver, Canada. She also has a Bachelors in Physics with a double minor in Astronomy and Mathematics from Western Washington University in Bellingham, WA, which she earned in 2007.

Dr. Greenstreet is interested in populations of small bodies that undergo unique dynamical phenomena, such as orbiting the Sun on backwards (retrograde) orbits or becoming trapped in orbits around the Sun very near the orbit of a planet where the asteroid and planet appear to “co-orbit” the Sun. Greenstreet also studies the rate at which small bodies in the Solar System impact planets, creating impact craters on their surfaces. Dr. Greenstreet has also used telescopic observations to help refine the orbits of newly discovered near-Earth asteroids, determine physical properties of asteroids of interest to NASA and upcoming targets of the Goldstone and Arecibo radar telescopes, and confirm that predicted asteroids undergo the Yarkovsky effect.

Recently, Greenstreet co-authored a paper with scientists on the New Horizons spacecraft science team to make predictions about the expected number and distribution of impact craters on the surface of the New Horizons flyby target, 2014 MU69 (a.k.a., Ultima Thule), which it flew passed on January 1, 2019. The paper was submitted for publication two weeks before the flyby and published in the Astrophysical Journal in early February 2019. The paper predicted a scarcity of craters on 2014 MU69 due to the number and distribution of craters observed by the New Horizons spacecraft on Pluto and Charon during its fly-through of the Pluto system in July 2015, which was published in Science magazine in March 2019 and co-authored by Greenstreet. Images recently sent back from the New Horizons spacecraft of 2014 MU69 indicate there are, in fact, few impact craters on the surface as Greenstreet’s paper predicted.

Dr. Greenstreeet is also interested in public outreach. She has given public lectures about her research and why asteroids are interesting and important to study at universities, high schools, elementary schools, and other events throughout the Pacific Northwest and southern California, including multiple appearances at the Santa Barbara chapter of Astronomy on Tap, where astronomers give short talks about interesting astronomy research in a bar setting. Greenstreet says, “Talking about the cool research we get to do as astronomers and explaining how scientists do research is, to me, one of the most important parts of the job. We all have a responsibility to share what we learn with the public to keep interest in science fields alive and to inspire the next generation of scientists who will take our research to the next level.”

Learn more about Dr. Greenstreet’s work at www.sarahgreenstreet.com

DiRAC Astronomers On Call: Chasing Gravity Wave Sources Using APO

The image shows the localization of the gravitational-wave (from the LIGO-Virgo 3-detector global network), gamma-ray (by the Fermi and INTEGRAL satellites) and optical (the Swope discovery image) signals from the transient event detected on the August 17th 2017 (Abbott et al. 2017) Image Close Up

The last prediction of Einstein’s general theory of relativity was the observation of gravitational waves emitted by two coalescing black holes, and yet again Einstein was proven to be right. Thanks to LIGO/Virgo instruments, we can now test General Relativity in the strong field regime using the gravitational-wave signals from merging black holes.

The first-ever direct detection of the gravitational wave (GW) signal, GW 150914, was made by Advanced-LIGO in 2015 September from a binary black hole (BBH) merger. This discovery entered us into the GW era, and was dramatically followed by the first observation of a binary neutron star merger, GW 170817, associated with a short gamma-ray burst (sGRB), GRB 170817A, inaugurating the era of multi-messenger astronomy.

DiRAC researches — Dr. Golkhou, Prof. Bellm, and Dr. Graham — are engaged in finding and characterizing optical counterparts to neutron star mergers (double neutron star or neutron star-black hole) discovered by the LIGO/Virgo Collaboration using the Zwicky Transient Facility (ZTF). ZTF is a new optical time-domain survey that uses a powerful new camera with a 47-square-degree field of view to find exotic and fast-evolving transients. ZTF and other facilities around the world are conducting Target of Opportunity observations of LIGO detections under the coordination of the Global Relay of Observatories Watching Transients Happen (GROWTH) collaboration. ZTF plans to trigger on all LIGO/Virgo triggers with a significant probability of having at least one neutron star in the system and a localization that is at least partially accessible from the Palomar night sky. In addition to ZTF, as a new member of the GROWTH project we have now access to more than 18 telescopes around the world which all have ongoing multi-wavelength follow-up programs and are actively searching for candidate EM counterparts to LIGO/Virgo sources. In order to fulfill our duty to the project, we brought the unique imaging and spectroscopic capabilities of the 3.5m telescope at Apache Point Observatory, New Mexico into the GW follow-up efforts. 

The goal is to discover compact binary coalescence (CBC) events — powerful engines for the production of GW, EM, and neutrino radiation. Rapid follow-up observations of such events are required to provide us a more accurate measurement of basic astrophysical properties such as the luminosity and energetics of this strong-field gravity event. Joint EM-GW detections constrain fundamental physical properties of CBC events including the distance scale, luminosity, and host galaxy environment. However, identifying a counterpart is remarkably challenging due to the LIGO/Vigo inherently weak localization of GW events (∼ a few hundred deg2). Nonetheless, the scientific returns of such discovery — and equally, event characterization — justify many efforts taken, even a small step forward.

Finally, after a long-period upgrading the LIGO (L1 and H1) and Virgo instruments,  the third observing run (O3) for detection of gravitational waves was began on April 1st, 2019. This new search would reach deeper into the universe (LIGO: up to ~170Mpc and Virgo: up to ~85Mpc for binary neutron-stars; BNS) than the previous two searches with the advanced detectors.

At the time of writing we have received thirteen triggers from LIGO/Virgo collaboration; the first one is S190408an and the last one S190521r. Three out of these thirteen triggers (S190425z, S190426c, and S190510g) are estimated to be associated with coalescence of BNS with a very high probability. Nine of these triggers are associated with BBH merger candidates and the remaining one could be the highly awaited NS-BH merger event, which has not been observed yet.

DiRAC astronomers as part of the bigger GROWTH collaboration team were actively engaged in the follow-up of candidates localized with the ZTF and DECam searches and suspected to be sources of BNS-merger signals detected with the LIGO/Virgo detectors. Though, no EM counterpart to those LIGO/Virgo BNS events has been identified yet. DIRAC Researcher and Moore-Sloan Postdoctoral Fellow, V. Zach Golkhou, is a coauthor on a recent paper regarding follow-up of a distant BNS merger candidate associated with S190510g. (https://arxiv.org/abs/1906.00806)

Figure 1 shows the localization of the GW (from the LIGO/Virgo 3-detector global network), gamma-ray (by the Fermi and INTEGRAL satellites) and optical (the Swope discovery image) signals from the transient event detected on the 17th of August, 2017 (image credit: Abbott et al. 2017, ApJL, 848, 12A). An artist’s impression of two stars orbiting each other and progressing (from left to right) to merger with resulting gravitational waves is also illustrated in the figure (image credit: NASA). It includes examples of the optical-infrared photometric and spectroscopic follow-up of GW 170817 from the GROWTH network, and other APO-class telescopes, as an example of the data that our program aims to obtain (images credit: Kasliwal et al. 2017, Science, 358, 6370 and Smartt et al. 2017, Nature, 551, 75).

So far, LIGO/Virgo have identified, on average, one BNS candidate every two weeks (two CBC events per week). This rate would exceed the expected value of LIGO/Virgo’s BNS identifications over the course of O3; 10 BNS/year. At this rate, many Astronomers including DiRAC-GROWTH team will be On Call all the time and probably are not going to get much sleep for a year while O3 is on!

Image Close Up


Astronomical Extensions for Spark (AXS) allows cross-matching multi-billion catalogs

Modern sky surveys are producing astronomical catalogs with billions of stars and galaxies. What is often important for science is cross-correlating these catalogs and finding the matching objects in several catalogs so that new insights can be gained from all observations at once. This operation, commonly known as ‘cross-matching’, can be extremely computationaly expensive because of the large number of comparisons that need to be performed.

DIRAC team has designed and implemented a system called AXS, or Astronomical Extensions for Spark, that comprises a new cross-matching approach that significantly outperforms other such systems and is capable of cross-matching multi-billion catalogs in tens of seconds on commodity hardware. AXS also contains other functionalities useful to astronomers and is based on Apache Spark, an industry-standard, open-sourced, distributed data processing system.”

Insights from MU69’s (Lack of) Craters

Recently, a team of scientists led by Dr. Sarah Greenstreet (B612 Asteroid Institute and University of Washington) conducted a study in which they made predictions for the crater count they expected to find on MU69’s surface. Dr. Greenstreet and collaborators used observations of Pluto and Charon’s surfaces and models of known Kuiper-belt populations to explore the bombardment of MU69 over the solar system’s life span and calculate the number of craters of different sizes its surface should host.

The authors’ results were intriguing: they found that, despite getting bombarded for 4+ billion years, MU69 should be marred by very few craters. Dr. Greenstreet and collaborators estimate that MU69 should have only ~25–50 craters larger than ~200 m in size, which is the smallest size we’re likely be able to see with the full-resolution New Horizons images.

Read full article provided by aasnova.org here.

DiRAC Scientists at the Asteroid Day 2019

During the last week of June, Asteroid Day celebrated their Fifth Anniversary in 2019, with events in 192 countries, and once again broadcasted their six-hour Asteroid Day LIVE TV show from Luxembourg, hosting innovators, astronauts, planetary scientists, celebrities and asteroid experts.

Scientists from the DiRAC Institute and LSST, Dr. Lynne Jones and Prof. Mario Jurić, were among the panel guests. The program featured two short films about team’s work and current research. “2017 The LSST – Making Census of the Solar System” and “2019 New Era of Cosmic Discovery“.

“Asteroids” Trailer from Nikolina Horvat on Vimeo.

New Era of Cosmic Discovery from Nikolina Horvat on Vimeo.