1 of 15

1

Welcome!

RRB I. AGN and Microlensing

Tue Aug 11, 10:30am-11:30am

Chairs: Franz Bauer and Sophie Reed

Slack: #day2-tue-slot2i-rrb-agn-and-microlensing

Agenda (30 min)

Welcome, Introduction, and Reminders (5 min)

Flash Talks (3 min each)

  1. Feature Definition and Classification of Microlensing Light curves, Somayeh Khakpash
  2. Hunting gravitational wave black holes with microlensing, Natasha Abrams
  3. XMM-SERVS: A Large XMM-Newton X-ray Survey of the LSST Deep-Drilling Fields, William Brandt
  4. Time domain mapping of light echoes from the core of AGN, Dragana Ilic

Question & Answer (10 min)

Attendees: Please mute!

​

For the Q&A, ask questions via the Zoom chat or the Slack channel, and wait for the chair to invite you to unmute.

​

We apologize if there is not enough time for all questions; please continue discussions via Slack!

Vera C. Rubin Observatory | Project & Community Workshop | August 10 - 14, 2020

#rubin2020

2 of 15

Friendly Reminders

Vera C. Rubin Observatory | Project & Community Workshop | August 10 - 14, 2020

#rubin2020

3 of 15

Friendly Reminders

Vera C. Rubin Observatory | Project & Community Workshop | August 10 - 14, 2020

#rubin2020

4 of 15

Somayeh Khakpash

  • Complex features in binary-lens microlensing light curves do not usually have direct relation to the physical parameters of the binary systems.
  • Simple heuristic functional fits can be applied to the light curves to describe these features.

​

Feature Definition and Classification of Microlensing Light curves

A schematic representation of a binary-lens microlensing light curve and its related features.

A heuristic model fitted to a binary-lens microlensing light curve.

Khakpash et. al. 2020 in prep

Khakpash et. al. 2019

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

5 of 15

Somayeh Khakpash

  • The resulting parameters from the heuristic functional fits are evaluated in different ways.
  • These parameters can be used as input features to machine learning classifiers.
  • Some of these algorithms are also useful for detecting microlensing light curves among other types of stellar variability.

​

An example of evaluating results of a heuristic function fitted to the light curves; the distribution of an estimated parameter for each class of lightcurve.

Khakpash et. al. 2020 in prep

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

6 of 15

Somayeh Khakpash

  • We have tested this method on high-cadence light curves simulated for Roman space telescope.
  • Preliminary results of training a few machine learning classifiers using the features extracted from the light curves seem to work successfully.
  • We are interested to investigate to what extend this approach is applicable to Rubin microlensing light curves and which observing strategies would yield the most optimum results.

​

Using Neural networks to classify into stellar binary-lens and planetary binary-lens microlensing light curves

Using the Random Forest classifier to classify into single-lens and binary-lens microlensing light curves

Khakpash et. al. 2020 in prep

Preliminary results of testing machine learning classifiers suggest this approach is successful.

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

7 of 15

Natasha Abrams

Hunting gravitational wave black holes with microlensing

​

Microlensing Event Rate Summary

​

  • Depends on the spatial, velocity, and mass distribution of lenses
  • Velocity distribution assumed to be the same regardless of component (white dwarf, black hole, etc)
  • Ends up with an M^2 dependence
  • Challenge would be to find the mass distribution from event rate

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

8 of 15

Natasha Abrams

  • Same number of BHs in both models
  • Due to M^2 dependence, higher mass BHs lead to large shoulder

​

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

9 of 15

Natasha Abrams

  • Simulate 12,000 lightcurves with various umin, t0, and tE
  • Sampled using candidate LSST cadences
  • Two distinct criteria: triggers and detections
  • Trigger Criteria:
    • 10% amplification (at least 2 observations)
  • Detection Criteria:
    • 10% amplification
    • At least 4 observations on either side of the peak
  • altLike > bulge > baseline
  • Excluding edge effects, at large tE there is high detection efficiency regardless of cadence

​

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

10 of 15

W. Niel Brandt

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

11 of 15

W. Niel Brandt

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

12 of 15

W. Niel Brandt

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

13 of 15

Dragana Ilic

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

14 of 15

Dragana Ilic

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020

15 of 15

Dragana Ilic

Vera C. Rubin Observatory Project and Community Workshop 2020 | August 10 - 14, 2020

#rubin2020