SLIDE 1
The curious DM events of J1713+0747
Fang Xi Lin w/ Michael Lam, Hsiu-Hsien Lin, Jing Luo, Ue-Li Pen, Dana Simard
CITA, University of Toronto Data from NANOGrav Collab.
Intro to corrugated sheets
Pen & Levin 2014 Simard & Pen 2018
The curious DM events of J1713+0747 Fang Xi Lin w/ Michael Lam, - - PowerPoint PPT Presentation
The curious DM events of J1713+0747 Fang Xi Lin w/ Michael Lam, - - PowerPoint PPT Presentation
The curious DM events of J1713+0747 Fang Xi Lin w/ Michael Lam, Hsiu-Hsien Lin, Jing Luo, Ue-Li Pen, Dana Simard CITA, University of Toronto Data from NANOGrav Collab. Intro to corrugated sheets Pen & Levin 2014 Simard & Pen 2018
SLIDE 2
SLIDE 3
Intro to corrugated sheets
Simard & Pen 2018
SLIDE 4
NANOGrav 12.5y DM timeseries of J1713+0747
SLIDE 5
NANOGrav 12.5y DM timeseries of J1713+0747
SLIDE 6
NANOGrav 12.5y DM timeseries of J1713+0747
SLIDE 7
NANOGrav 12.5y DM timeseries of J1713+0747
SLIDE 8
Fitting a thin, underdone crest
1st event χ2
reduced = 1.81
SLIDE 9
Fitting a thin, underdone crest
1st event 2nd event χ2
reduced = 1.81
χ2
reduced = 0.99
SLIDE 10
Underdense crest predictions
- 2 images before crossing the
crest: line-of-sight & lensed image
up to 40% of
- 1 image after crossing the
crest
μlensed μLoS
SLIDE 11
Underdense crest predictions
- If you time-resolve the images,
you should see an echo.
- If you don’t, they could
interfere.
SLIDE 12
Underdense crest predictions
- For an observation ~ 1 month before
the 1st event,
- For
, we expect an interference frequency width of few hundred kHz, below NANOGrav’s 1.5MHz frequency resolution.
- A closer screen would have a larger
frequency width, that Michael Lam & NANOGrav collaborators would have detected.
- A farther screen would lead to a
dimmer lensed image, but larger delay profile change (pre-crossing)
dlens = dpsr/2 ∼ ⇒
SLIDE 13
SVD Analysis
- “Eigendecomposition” of data matrix.
- Tailor-made to analyze whether there are profile variations
from observation to observation.
- We SVD 11.5 years of profiles (separately for each
telescope at each observing band), and project out the first two modes.
SLIDE 14
Greenbank, 1.5GHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 15
+ Greenbank, 820MHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 16
+ Arecibo, 1.4GHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 17
+ Arecibo, 2.0GHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 18
+ Greenbank, 1.5GHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 19
+ Greenbank, 820MHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 20
+ Arecibo, 2.0GHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 21
+ Arecibo, 1.4GHz
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 22
Unlikely to be instrumental effect!
MJD MJD Pulse phase Pulse phase Pulse phase Pulse phase
SLIDE 23
Profile zoom at Arecibo 2.0GHz of the second event
Normalized profiles & SVD residuals, pre-dip
SLIDE 24
Profile zoom at Arecibo 2.0GHz of the second event
Normalized profiles & SVD residuals, post-dip
SLIDE 25
What else can it be?
SLIDE 26
What else could it be?
- Dedispersed to the wrong DM after the the dips.
- There’s an echo, but after the dips.
- There’s some extra scattering after the dips.
- Extra RM and/or leakage to Stokes I
SLIDE 27
Wrong DM?
SLIDE 28
Echo?
SLIDE 29
Extra scattering?
SLIDE 30
Polarization?
SLIDE 31
Summary
- J1713+0747 crosses 2 events that looks like underdense
crests of a corrugated sheet in 12.5 years.
- We do not see evidence for a predicted second image
pre-crossing, but perhaps we can with better data.
- We do see profile shape change post-crossing, which is
unexpected, and maybe unexplained by an underdense crest.