Neronov et al 2015: microlensing in PKS1830-211

[drphilmarshall]

Hi all,

I just finished reading this paper, and we'll discuss it further at KIPAC tea this morning. They argue persuasively for a model where the gamma rays come from a small (10^14 - 10^15 cm) emission region close to the black hole, where the size constraint comes from the range of magnification ratios seen between pairs of time-delayed flares. They see two features in the structure function (similar to the ACF), which they interpret as a gravitational lens time delay of 26 days consistent with the value found in the radio, and a longer lag at 76 days that they suggest might be a quasi-periodic flare repeat timescale.

This means that the TimeBombs model is close to what they would assume, except that they would allow the magnification ratio to be different for each flare! Indeed, they claim that the strongest flares are associated with microlensing caustic crossing events, which would have Pacynski-style profiles instead of our spikier flares...

[drphilmarshall]

I spoke to Krzyzstof Nalew (@knalew) after we discussed the Neronov PKS1830 paper at tea yesterday. He is skeptical about their claim to be able to rule out the jet irregularities model (where compact emission regions in the jet, ~pc away from the central engine, are the sources of the flares). We noted that microlensing would be present in both this model and the Neronov central source model, with source size needing to be taken into account to predict the microlensing strengths accurately. The jet model has the additional (superluminal!) speed of the source to be taken into account though. We agreed that training the underlying flare model on a sample of unlensed blazars was a Good Thing, so we can assign informative priors to this part of the model before (potentially) doing model comparison between the different source models. Either way, with microlensing on the table, we'd need to be able to extend the TimeBombs model to allow a different magnification ratio for each flare, while preserving a single time delay - although Barnacka has the jet extending so far that the time delay should be different as well (and correlated with the macrolens magnification...) Would this be easy to do, @eggplantbren? Give each flare its own microlensing magnification, as well as the global macrolensing magnification?

[drphilmarshall]

Meanwhile, we have the energy-dependent pair creation effect to consider as well: this ought to be modeled as part of the source size as well! @knalew, can you comment on what a good spatial model for a gamma ray emission region might be, please? A 2D circularly-symmetric Gaussian with size parameter sigma, where sigma changes in some way (a power law?) with energy band? Interested to hear your thoughts on possible parametrizations and associated priors, for both the central and jet models!

[knalew]

When modeling the activity of blazars, we usually think of 'blobs' or 'shells' propagating along the jet. A typical approach is to assume a certain function for injection of energetic electrons, e.g. a flat injection profile between some r0 and 2 r0. The electrons evolve mainly due to radiative cooling and adiabatic losses (due to jet expansion). There are more sophisticated approaches including stochastic acceleration, particle escape, etc. The appearance of the gamma-ray emitting region was never of much interest, although I did play with that a little. A circular Gaussian profile should be fine to begin with, its size may be allowed to increase gradually together with the expanding jet. The critical question is the actual lifetime of a flaring event. The apparent superluminal motion of the gamma-ray emitting region needs to be accounted for. Apparent motions up to 20-40c are routinely observed in blazars with radio/mm VLBI. It means that for individual flares the microlensing caustics can be considered stationary. Another potential twist is that consecutive 'blobs' do not necessarily follow exactly the same trajectory. Radio knots often show a lateral spread and velocity components perpendicular to the jet. This could also account for some variations in the magnification ratio between the flares. The energy-dependent effects and the gamma-ray opacity are additional complications, but I would leave them for now. I hope this is helpful.

[drphilmarshall]

OK, I think this is what Neronov et al are getting at: out in the jet, the caustic crossing timescale should be much shorter and so we should see narrower microlensing spikes on top of the broader intrinsic flares. In the unlensed blazars we should see no such narrow flares. The Neronov et al claim is that there is additional microlensing magnification on timescales comparable to those of the underlying flares.

Regarding energy dependence: this could be an important extra piece of observational evidence that has not yet been used. At the very least we shoudl find consistent time delays from one band to the next, but if we see different microlensing magnifications that could be telling us a little more about the source regions.

On Wed, Jul 8, 2015 at 10:06 AM, knalew notifications@github.com wrote:

When modeling the activity of blazars, we usually think of 'blobs' or 'shells' propagating along the jet. A typical approach is to assume a certain function for injection of energetic electrons, e.g. a flat injection profile between some r0 and 2 r0. The electrons evolve mainly due to radiative cooling and adiabatic losses (due to jet expansion). There are more sophisticated approaches including stochastic acceleration, particle escape, etc. The appearance of the gamma-ray emitting region was never of much interest, although I did play with that a little. A circular Gaussian profile should be fine to begin with, its size may be allowed to increase gradually together with the expanding jet. The critical question is the actual lifetime of a flaring event. The apparent superluminal motion of the gamma-ray emitting region needs to be accounted for. Apparent motions up to 20-40c are routinely observed in blazars with radio/mm VLBI. It means that for individual flares the microlensing c austics can be considered stationary. Another potential twist is that consecutive 'blobs' do not necessarily follow exactly the same trajectory. Radio knots often show a lateral spread and velocity components perpendicular to the jet. This could also account for some variations in the magnification ratio between the flares. The energy-dependent effects and the gamma-ray opacity are additional complications, but I would leave them for now. I hope this is helpful.

— Reply to this email directly or view it on GitHub https://github.com/eggplantbren/TimeBombs/issues/12#issuecomment-119665276 .