AY3: Freshman Seminar (Automated Discovery of the Universe)
Astronomy is now and will continue to undergo a profound change.
When I was a graduate student (which, I remind you, was in the last
century) astronomy was a data starved area of science. It took great
effort to simply collect the data. The analysis tools were accordingly
suited to small amounts of data. In contrast, today, the field of
optical/IR (OIR) astronomy is inundated with data -- thanks to
technological progress which has given us large format detectors,
decreasing costs for unit telescopes, advances in electro-mechanical
systems (allowing robotocizatoin) and of course all the advances
in computers and related areas. A similar revolution is happening
in radio astronomy.
Who will succeed and make discoveries in the new era?
Answer: those who will have ability to deal with big data sets,
connect diverse catalogs and invent new algorithms whilst being
grounded in astronomical phenomenology and physics.
The purpose of this class is to give you a taste of the new era.
The class will revolve around the Zwicky Transient Facility (first
light is July 2017), the successor to the recently concluded Palomar
Transient Facility (PTF). You will be introduced to the methodology
of time domain astronomy and undertake several pedagogical exercises
involving periodogram, FFTs and statistical inference. You will
then have an opportunity to analyze fresh data. Should you
make a discovery, even if modest, then you will be encouraged to
write up as a real paper and submit to a journal. Along the way you
will be asked to give presentations of your work (and thus help
develop good speaking skills, an essential part of life).
The class meets in Cahill, room 211 as follows: Monday 2-3p and
Friday 3-4p. The timetable below presented below is the planned
timetable. It will get updated as we progress along the course.
Data analysis requires strong skills in programming and also an
ability to be a "power user". It is my view that a good researcher
must be deeply ground in UNIX and Python (and Matlab for specialized
research, especially in applied sciences & engineering). I would,
in fact, say that a key benefit of taking this class is that you
will get a good start along these lines.
- Please bring your computers to the class
- Please have UNIX/LINUX or appropriate Windows/UNIX interface loaded up.
- Please have Python loaded up
- Introduction to Time Domain Astronomy (TDA) [6 Jan]
Talk by Prof. Hillenbrand.
Top Argelander Stars: Pedagogy & Prize
- The Zwicky Transient Facility (ZTF) [9 Jan]
Talk by Dr. L. Yan & lab tour by Roger Smith
- The Gaia Mission
Talk by Dr. N. Blagorodnova. Presentation
- Time & Calendar [13 Jan]
Solar day, Cesium clock second, Gregorian calendar, Julian Day Number
- Astronomical Coordinates (RA, DEC) [16 Jan]
Altitude, Azimuth, Right Ascension, Declination
- Local Sidereal Time & Offsets [ 18 Jan]
Sidereal time, LST at midnight, Computing angular offsets
(spherical trig; arcseconds)
- Photometry & Magnitude [20 Jan]
Poisson noise, Background, Aperture Photometry
Magnitude, Zero point (AB, Vega), Error on Magnitude
- Introduction to UNIX [20 Jan]
pwd, cd, mkdir, ls, cat, rm, mv, rmdir, more, less, head, tail,
which, info, man, wc, wget, curl, compress, uncompress, paste,
grep, sed, [.,*,..,#,|,~,$] cut, sort, uniq, tr, df, du
perhaps the most complex Unix command
- Simple exercise in Statistics [23 Jan]
I reviewed basic concepts in statistics (mean, median, histogram) and also
first introduction to reading files (UNIX, Python etc). See
- Introduction to Python [Jan 25]
Nadia's Python presentation
|  Supernova lightcurve data file
- Period Folding [Jan 27]
This class is focused heavily on periodic signals. In optical astroomy
RR Lyrae are famous for distance indicators. These are pulsating stars.
Please review the presentation on RR
Next, please read the note.
Next, please review my MATLAB demonstration .
Please read the file PeriodicSignal.dat. Fold this
file at period clearly noted in the demonstration package. Use 21
bins for the folded profile. However, plot the profile for two periods
(i.e. plot the 21-bin profile twice). Why? Change the period to a
value and then fold the input data series.
- RR Lyrae from PTF [Jan 30]
ASCII files and fold time series to given period.
- Introduction to the Galactic Marshal [Feb 1]
- Fourier Transforms [Feb 3]
Introduction to Fourier Transform. The formal notes for this
specific class can be found here
- Visit & Overnight Stay at Palomar [Feb 4]
Tour of the grounds, P48, P60 and
- Understanding Fourier Transforms [Feb 6]
You should work through the tutorial.
- Application to Pulsar Astronomy [Feb 8]
Pulsar Data Base
please find time series for three pulsars. The headers are given
in ascii files. The data is simply a series of 32-bit "floats". Read
the data and Fourier transform the time series. Plot the power spectrum.
You will notice the low frequency spectrum is huge. Ignore this (i.e.
set the y-limits) and look for the harmonics. You may benefit from reading
Notes on FFT and
also reviewing my Matlab
Each of you is expected to present your analysis of the pulsar data (I suggest
that you choose either bright or medium pulsar, not the faint one).
Here are guidelines for presentation.
Please have your talk ready as a pdf file.
- Presentation by students (FFT/Pulsars) [Feb 10]
Rules for Giving Talk
- PTF data: An introduction to periodic stars [Feb 13]
Dr. Thomas Kupfer Presentation
- Periodogram: introduction and application to PTF data
Dr. Thomas Kupfer.
Data sets are provided.
Every dataset consists of about 20 stars. Most of them were selected
from PanSTARRS colors and show signs of variability in PTF but have not
been tested for periodic variability.
is expected to analyze one data set and report on Feb 22 (see below).
- The Kepler Mission [Feb 17]
- Student presentation of Light Curves of PTF data
- Student presentations continued [Feb 22]
- Discussion of Findings [Feb 24]
Studnets discuss the nature of binaries (tidal, reflection, eclipsing, Algol, ..).
Exciting report: discovery of a rare g-mode pulsating sdB/sdO star by student
- Discussion of Findings, continued [Feb 27]
Graduate student Kevin Burdge explains sample selection
and reports other new finds from
the larger PTF data set of other interesting binaries.
- Introduction to searching for pulsators & planets with Kepler [March 1]
Dr. Courtney Dressing will walk us through Kepler data analysis and conclude
with a talk on her research. The data sets can be found here
- In class laboratory: Anlysis of Kepler datasets led by Dr. Dressing
Students should work on the three notebooks in order given:
(1) Box Least Squares periodogram and search for
(2) BATMAN package to model transit light cuves with parametric input
and (3) overplot Kepler data and transit models
- Inference: Introduction [Mar 6]
Poisson process: time between intervals, Poisson statistics.
- Inference: Confidence Intervals [Mar 8]
Numerical experiments: what is the distribution of time between
gamma-ray bursts? For a measured number of photo-electrons determine
[1%, 99%] confidence interval.
- Inference: Normal Distribution [Mar 9]
The central role of Gaussian distribution. The Central limit theorem.
The concept of "3-sigma" and the occasional fallacy.