Waking from the Nightmare Scenario

[kadrlica]

According to Buckley and @ahgpeter, the nightmare scenario occurs when no dark matter signal is seen in any of the particle physics searches in the next decade. How can astrophysics wake us from the nightmare scenario? One key predictions of CDM is the existence of very low mass dark matter halos. We would like to get to 10^-6 Msun, but maybe we should start with something a bit more reasonable...

@bechtol and @kadrlica have been talking about a calculation that would provide a quantitative estimate of the (non?) detectability of low mass (< 10^5 Msun) Milky Way dark matter subhalos using correlated nanolensing signatures (i.e., the correlated photometric brightening in a collection of stars). While we are aware of several nanolensing estimates in current years, we are not aware of any trustworthy quantitative estimate of such a signature.

A set of toy parameters to define the problem:

• A ~10^4 Msun dark matter halo with a scale radius of ~50 pc located ~25 kpc from Earth in the direction of the LMC with >300 stars per arcmin^2 (lower estimate from Table 2 of Alcock et al. 2001).
• This estimate assumes that the radius of the dark matter halo scales as M^(1/3) and assumes that the dwarf galaxies have M = 10^9 Msun and rs = 1 kpc.
• 25 kpc is chosen to be half-way between Earth and the LMC.
• The projected radius of such a halo would be ~0.1 deg^2 ~ 400 arcmin^2 => 1.2e5 stars.

There are some know references from people in this group:

• https://arxiv.org/abs/0905.1998
• https://arxiv.org/abs/1008.2385
• https://arxiv.org/abs/1310.5412

[jobovy]

I'm interested in looking at this as well and was discussing something like this with @sbird earlier today as well. One positive comment: from Via Lactea halos we found that

r_s = 1.05 kpc (M/10^8 Msun)^0.5

in Bovy et al. 2017 and Erkal et al. 2016. So a 10^4 Msun halo would have rs ~10 pc rather than 50 pc; that should make this a little more detectable.

[kadrlica]

I think the possibility of a correlated nanolensing signal might be the kind of extremely difficult observational problem that piques the interest of @smurgia or @rmandelb.

[rmandelb]

Clearly Alex knows me well. As soon as you put nano in front of "lensing" it's simultaneously terrifying and intriguing to me. That could be a hack project for tomorrow morning if somebody wants to discuss with me? @jobovy @sbird or anybody else? We could go through

• the actual magnitude of the effect expected?
• an estimate of its observability (or not) given the LSST survey parameters?

I've not looked at the above references at all yet, so I'm not sure to what extent they answer the first question.

[kadrlica]

@rmandelb my (naive) understanding is that there has not been a cohesive estimate of the magnitude of this signal, but that it is almost certainly way below the systematic limits of LSST. I think what we are looking for as an estimate for some far-future, extremely stable space-based observations. @bechtol has looked into this a bit more and might be able to comment.

[rmandelb]

That makes sense. Either way, estimating the observational parameters needed sounds very interesting to me.

[bechtol]

There is a series of recent papers regarding a candidate extreme magnification event for an individual star passing behind a cluster caustic, e.g., https://arxiv.org/abs/1706.10279 . In this lens system, the star is magnified by a factor of several thousand! Most of the interpretation papers have focusing on constraints for compact DM models. However, given the extreme magnification that would substantially enlarge the effective Einstein radius, it seems that there might be sensitivity to DM substructures in “vanilla” CDM scenarios. The cluster caustic lens systems have other attractive properties in that the source is very compact, and in contrast to the multiply-lensed quasars, one does not need to compare images and fit time delays to back out the intrinsic source variability since stars are more stable in their flux. In other words, the observed variability is probably mostly due to lensing effects.

Along these lines, there was a paper a couple of years ago from Rahvar et al. 2013 (https://arxiv.org/abs/1310.5412) proposing that even if the signal from individual dark matter substructures is very small, the ensemble of numerous tiny halos crossing the line of sight might induce variability with a distinctive power-spectrum from stellar microlensing. They considered multiply imaged quasars in that paper, but it seems that a cluster caustic lens system could be more sensitive, for the reasons mentioned above.

Anyhow, this is a new observational phenomena published for the first time in 2017, and it seems that LSST would be well suited to identify additional cluster caustic lens systems. I wonder with an appropriate monitoring campaign with precise photometry if one could use such systems to constrain DM substructures at halo masses orders of magnitude smaller than our current observational probes. If LSST can find the cluster caustic lensing systems, maybe follow-up space-based monitoring (I know that sounds a bit crazy) could be used for more precise photometry.

The theory work is for someone to compute the lensing magnification signal as a function of time for an ensemble of CDM halos along the sightline, and whether this can be distinguished from stellar microlensing (cadence of observations and relative photometric precision requirements).

[kadrlica]

@bechtol Thanks for the interesting note. The observational signatures are different from the original proposal, but definitely in line with getting out of the nightmare scenario.

[bechtol]

I also posted a related comment here #9 regarding astrometric anomalies. In any case, it would be useful to code up the calculation of what the magnification and astrometric deflection would be as a function of subhalo mass, subhalo profile shape, subhalo distance, source distance, and angular impact parameter. I'm guessing this would not be too difficult for people who have previously thought more about lensing / there are standard tools for doing this.

[kadrlica]

Adding @chuckkeeton and @franyancr. I suggest that we merge #9 into this issue. What do you think @smurgia?

[smurgia]

@kadrlica I've been talking about #9 a bit more with people this morning and as that issue is shaping up, it might be best to keep it as a separate one. I'll add progress on the discussion on that soon.

[kadrlica]

A recent paper suggested by @nrgolovich: https://arxiv.org/abs/1804.01991
Also relavent to #9