Applications Of Computer Science in Astronomy

Applications Of Computer Science in Astronomy

• 1.
  Computers in Astronomy AhmedAbuzuraiq-ICS
• 2.
  Presentation Outline • AdaptiveOptics • Automated Observatories • Classification • Simulations • Final Words
• 3.
  Adaptive Optics • Weget twinkling or distorted images of celestial bodies because of atmosphere turbulence • Speckle imaging • Take multiple very short exposure images • (for optical light : 10 ms)then combine them. Before After Diff. in refractive index
• 4.
  Automated Ground Observatories •Too many things to observe and they have different priority. • E.g. Planets transits , eclipsing stars , Near Earth Objects , sun irregular activity , objects about to set ,..etc • Sometimes we would like to observe one body simultaneously from multiple observatories. • Previously, astronomers would meet and decide what is worthy to observe aided with a sky map • Takes time • Harder to synchronize and orchestrate with other observatories. • They have to handle how the schedule is going to change if there was an emergency (e.g. a bad weather or loss in communication.)
• 5.
  Cont. Automated GroundObservatories • Solutions : • Automate target selection • give a value for each target based on some parameters then choose target with highest value • Parameters are like : time to set , height in sky, how rare is the event proximity to the moon ,brightness ,…etc • A central control that automates synchronization • With experts intervention
• 6.
  Program Number PrincipalInvestigator Program Title 13665 Bjoern Benneke, California Institute of Technology Exploring the Diversity of Exoplanet Atmospheres in the Super-Earth Regime 13715 Jennifer Sokoloski, Columbia University in the City of New York Imaging Spectroscopy of the Gamma-Ray Nova V959 Mon 13740 Daniel Stern, Jet Propulsion Laboratory Clusters Around Radio-Loud AGN: Spectroscopy of Infrared- Selected Galaxy Clusters at z>1.4 13769 Klaus Werner, Eberhard Karls Universitat, Tubingen Trans-iron group elements in hot helium-rich white dwarfs 13794 John T. Clarke, Boston University Seasonal Dependence of the Escape of Water from the Martian Atmosphere 13845 Adam Muzzin, University of Cambridge Resolved H-alpha Maps of Star-forming Galaxies in Distant Clusters: Towards a Physical Model of Satellite Galaxy Quenching 13868 Dale D. Kocevski, Colby College Are Compton-Thick AGN the Missing Link Between Mergers and Black Hole Growth? 14057 Fabien Grise, Universite de Strasbourg I Changes in the X-ray irradiation of an ultraluminous X-ray source 14071 Sanchayeeta Borthakur, The Johns Hopkins University How are HI Disks Fed? Probing Condensation at the Disk- Halo Interface 14076 Boris T. Gaensicke, The University of Warwick An HST legacy ultraviolet spectroscopic survey of the 13pc white dwarf sample 14077 Boris T. Gaensicke, The University of Warwick The frequency and chemical composition of rocky planetary debris around young white dwarfs: Plugging the last gaps 14080 Anne Jaskot, Smith College LyC, Ly-alpha, and Low Ions in Green Peas: Diagnostics of Optical Depth, Geometry, and Outflows 14095 Gabriel Brammer, Space Telescope Science Institute - ESA Calibrating the Dusty Cosmos: Extinction Maps of Nearby Galaxies HST Programs: November 30 - December 6, 2015 In Hubble Space Telescope ,people make proposals for possible targets and an algorithm choses in which order to execute accepted proposals . Automated Observatories-HST
• 7.
  Classification • So wewant to classify objects in our images • What kind of galaxy is this ? • Is this a globular or open star cluster ? • … • But The amount of data collected is huge • Different wavelengths , new high resolution techs means more data . • Large Synoptic Survey Telescope (LSST). Planned to enter operation in 2022,is aiming to gather 30TB a night. • There is relatively a few number of astronomers (both professional and amateurs)
• 8.
  Classification: Zooniverse • “TheZooniverse is a collection of web-based Citizen Science projects that use the efforts of volunteers to help researchers deal with the flood of data that confronts them.” • You are given a short tutorial on what is the benefit and how to help. • Then you are given images and asked to do something or answer questions on them • Examples: • Disk Detective. (detect stars with hidden debris discs around them) • Planet Four: Terrains (mapping the terrains of Mars) • Asteroid Zoo • Planet Hunters (exoplanets finding through transit method) • Sunspotter. • Spacewraps. (on gravitational lensing) • … • And others in both astronomy and other fields.
• 9.
  Classification : A.I.and Machine Learning So we have images of galaxies and from what volunteers did in Zooniverse We know the galaxy type in each image . Volunteers are asked questions one after the other starting from the top . ------->
• 10.
  Classification : A.I.and Machine Learning • So we train the program on this knowledge and it can then use this to classify galaxies that it hadn’t seen before. • Much like how you could teach child by letting him have many experiences ,he then will use that if he faced a new situation. The results were almost similar to the volunteers response !
• 11.
  Simulations • Some phenomena'scan only be studied through simulations • e.g. stellar collisions ,galaxies interaction • Also we can confirm our hypothesis using simulations • So we observe , make hypothesis ,make predictions and prove them. • What to simulate ? • We want to capture all the laws that we think are necessary. • So speaking of asteroids gravitational forces is enough , but when talking about stellar evolution we need to add hydrodynamics too . • We can use simplifying assumptions • Treat bodies as points • Assume bodies don’t collide • Assume the sun is stationary and is not wobbling. • If our simulation spans over a small scale we don’t usually need to account for what happens at larger scales.(and vice versa) , unless there is a significant relation .
• 12.
  Example : N-BodyProblem • Problem description • The n-body problem is the problem of predicting the individual motions of a group of celestial objects interacting with each other gravitationally. • 2–body and a restricted form of 3-body are the only ones solved analytically (i.e. there is a closed form ,a formula) 2-Body 3-Body
• 13.
  • first thingwe do is to set our initial conditions (position and velocity of each body). • Then we compute the acceleration. • Next step after finding the acceleration is to use it (the following is really a numerical intergration) (dt is the time step) • x += vx*dt (first we update position based on old velocity) • y += vy*dt • z += vz*dt • vx += ax*dt (then we update the velocity based on new accelation) • vy += ax*dt • vz += az*dt • What about when n > 3 : • Numerical methods must be used. Cont. N-Body Problem
• 14.
  Nbody Simulation atwork Initially Stationary - > clustering Non-zero initial angular momentum -> satellites form
• 15.
  Application : NiceModel • Nice Model : is a model of the early evolution of the Solar System. • simulations of the first Nice Model and the modified one (which have different assumptions ) can show us what are the consequences of this change in assumptions. Nice 1 at different times (notice Uranus and Neptune orbits ) Difference between the two models Is in how they interpret this change of orbits
• 16.
  Final words • Theapplications of computer science are everywhere . • That’s because computer science deals with the algorithmic side of mathematics . • So when mathematics tells you what is a square root ,computer science tells you how to compute it efficiently. • So….Keep a keen eye for the problems you face that might be solved by computers.
• 17.
  • “Computer Scienceis no more about computers than astronomy is about telescopes” Edgar Dijkstra • It’s rather really about the way of thinking.
• 18.
  Thanks for Listening!