The Universe extends above us in a sea of questions. Billions of galaxies, each with billions of stars and unique planetary systems, pose questions of origins and formation: the origins of the first stars and the planets, the formation of structures within the universe, and the beginnings of life. As the celestial scene moves and changes in layers of pulsing light, flashing with the births and deaths of stars, with the transience of bodies blinking through the cosmos, we undertake a quest to unravel this complexity, to make sense of this vast and changing Universe. For in this expanse, there is order and grandeur. There are signs of our long past and portents of our future. And there are answers to some of humanity's longest-held questions.
Now more than ever before, we are on the brink of an era of unparalleled cosmic discovery. Over the next decade, a new generation of telescopes and astronomical surveys will yield a 1000-fold increase in the amount of available astronomical data.
Each of these experiments holds great promise for the discovery of new and fundamental physics, ranging from characterizing the nature of dark energy and the composition of dark matter, to discovering the most energetic events in the universe, to tracking asteroids whose orbits may intersect with Earth.
But with this impending deluge of data, the scientific community will almost immediately find itself unable to process and analyze datasets of this scale and complexity in any meaningful way. Our potential for scientific discovery will no longer be limited by how we collect or store data – how much we can observe and catalog – but rather how we interact with and analyze these data. As a result, the very usefulness of these petabytes of astronomical data will hinge on what we do now to prepare.
At this critical juncture, the field of astrophysics must change the way it conducts science. The proliferation of powerful telescopes is a testament to our capacity to build cutting-edge instruments and collect astronomical data. But we must go beyond building and collecting. The new generation of telescopes and surveys requires new algorithms and scalable frameworks to deal with datasets of a complexity and scale that defy the standard research tools in our field. The scientific community is simply not prepared to maximize the potential of the data it will possess in a few short years.
We will soon reach a point at which software is the chief instrument for exploring the universe. As such, we must move toward a new model of data intensive, computationally-driven science – one that is profoundly interdisciplinary, uniting computer scientists, statisticians, astrophysicists and cosmologists to develop the computational solutions to problems presented by massive data streams.
If we do not move in the direction of collaborative, team-based science to develop new tools of data analysis, the billions of dollars invested in powerful instruments and ambitious surveys like the LSST will be wasted. Worse still, we will miss out on discoveries that will fundamentally change the way we understand the universe.
By the time data is being taken from the aforementioned surveys and the need for science driven algorithms is broadly recognized, it will be too late. The scientific impact of the programs will already be diminished.
Research institutions around the world face this pressing challenge, and the stage is set for an institution to take the lead in pioneering a new model of astrophysical science that anticipates and responds to the data rich landscape of the coming decade.
The University of Washington is positioned to emerge as the leader in innovative astronomical research and resource development through the Center for Data Intensive Research in Astrophysics and Cosmology.
The center will bring together researchers from Astrophysics, Computer Science & Engineering, Statistics, and other key disciplines, building on the success of the University of Washington’s eScience Institute. This coalescence of top UW researchers across the sciences will catalyze the development of algorithms and infrastructure to meet the needs arising from the major astronomical imaging telescopes (LSST, MWA, SKA, Advanced LIGO, and others).
In this collaborative environment, focused research groups composed of postdoctoral fellows and data scientists, under the leadership of UW faculty from relevant disciplines, will focus initially on two pressing astrophysical and cosmological research areas:
Driven by the availability of the new astrophysical surveys, these teams will develop new approaches for managing, storing, and accessing petabyte data sets and running analyses at scale and within these databases. This work will provide models for handling massive datasets while likely advancing scientific understanding on fundamental topics: The origins of the first stars, the rate at which the universe expands and its mass and energy content, the formation of planetary systems, and the location and trajectory of asteroids – including those on a collision course with Earth.
We believe that the strategic grouping of postdocs (with science and computational knowledge), proven data and software experts, and faculty leaders from the across the UW will accelerate algorithm development, provide evaluation of advances and breakthroughs in the context of non-astronomical science, and enable additional collaborations with researchers across campus. In the initial years of operations beginning in 2016, we expect to:
As the scientific community awaits an influx of astronomical data, the UW has the distinct opportunity to be the first to develop the tools and practices to handle that data. More than any other research institution in the world, the UW has the ideal mix of faculty, a clear vision, a track record for interdisciplinary collaboration, and strong connections to the incoming datasets.
In fields ranging from computational astrophysics to computer science to particle phenomenology, our researchers work at the interface between complex data and cutting-edge science. From the first digital survey of the sky (the Sloan Digital Sky Survey) to the highest priority for ground-based astronomy for the next decade (the Large Synoptic Survey Telescope) the University of Washington has taken a leadership role in survey astronomy. We are currently home to the leads in LSST’s Project Science (Željko Ivezić), Data Management (Mario Jurić), and Simulations (Andy Connolly); all of whom will be faculty leaders in the center, along with confirmed faculty from Computer Science & Engineering and Statistics.
No other institution in the world has a stronger connection to the incoming data coupled with strategically positioned faculty capable of transforming the way that data is managed and analyzed.
The center represents the culmination of ten years of crosscutting research, strategic hiring, interdisciplinary partnerships, and proven collaborative relationships with technology companies like Microsoft and Google. It provides a remarkable opportunity to change the way we undertake astrophysics and to answer some of the most fundamental questions about the origins of our Universe.