EC comparisons with JetSet: BLR

[cosimoNigro]

Hello,

let us discuss in this thread the comparisons for EC with jetset that were suggested by Justin. I started to work on them in the jetset_ec_test branch. Here let us comment on the BLR.
Here the link to the jetset documentation on External Compton. I think there is a 1:1 correspondence with the parameters we use, see the test_jetset_ec_blr.py script in the branch. The shell cannot be made infinitesimally thin, so I set the outer radius to be 1.01 times the inner one.

I used the same distances of the crosscheck:

• close to the BLR we have a similar shape but an order of magnitude difference

• far from the BLR we have two orders of magnitude discrepancy close to the peak

Let me know what do you think.

[cosimoNigro]

I made the BLR slightly narrower and used this feature that allows to transform the electron distribution in the disk frame (the docs say it should be used for distances smaller than the BLR radius).

The jetset SEDs is very near to a factor 2 that produced by agnpy :man_shrugging:

[jsitarek]

Hi @cosimoNigro,

did you check how jetset is treating the angles? If you are within the BLR radiation field the angular distribution of the radiation does not matter much, but if you are far from it, it make a lot of difference if the photons are considered to be moving on axis, or at an angle to it.

[cosimoNigro]

Performing some of the tests @jsitarek requested in another thread:

before giving up with jetset, could you make two simple test:

1. compare the two with the case of non-relativistic blob (e.g. Gamma = 1.001)

Here goes the non-relativistic case:

• inside the BLR
• outside the BLR

2. compare the two forcing the observation angle to be on axis (theta << 1/ Gamma).

Here goes the on-axis case:

• inside the BLR
• outside the BLR

Will add the same tests for EC on DT tomorrow.

Scripts to perform these tests with jetset are now available in the master, see the jetset_tests repository.

[jsitarek]

Hi @cosimoNigro one more thing came to my mind. In the "far" case the distance is about 3 orders of magnitude larger than the size of the sphere. This is very tricky to compute (and also pretty unrealistic scenario), and the angles there could make a huge difference (there will be a difference between 0 deg (on axis) and 0.05 deg. Depending how the integrals are being done this can also end up with large uncertainties(if the integration is over the sphere of BLR this is more or less fine, but if the integration is done over the angles w.r.t. observation direction this can have problems here). Have you checked also the case of moderately distant source (d = a few - 10 * R_BLR)?

[jsitarek]

follow up from the e-mail discussion: after fixing the transformation issue in the jetset the BLR comparison looks quite fine (some differences at very large distances can be related to the accuracy of integrals). Such a comparison might actually be added to the paper. What do you think @cosimoNigro ?

[cosimoNigro]

After some work by @andreatramacere in jetset the comparison improved a lot:

And @jsitarek commented in the PR adding this comparison to the main code of the repository that:

the difference are way too large for confort, reaching a factor 2, it would be good to understand if they come from different assumptions. Otherwise, a possibility is that they are affected by the numerical integration uncertainties, do you have a possibility in jetset to apply a finer integration to see if the results do not change considerably? In particular the EC on BLR for r >> Rline difference might be related to the accuracy of integration over angles, those are really tricky cases

I think it definitely comes from different assumptions. Rather than a monochromatic line, the BLR in jetset is reflecting the entire disk emission, so I think is producing another black body, correct @andreatramacere? Can you point us to the reference you are using, if there is one, to model the BLR?

I don't think it's an accuracy problem since in the last PR we have increased the Lorentz factor and solid angle integration grids.

[andreatramacere]

@cosimoNigro, these are the references:

• For the 'disk' transformation: M. Georganopoulos, J. G. Kirk, and A. Mastichiadis. The Beaming Pattern and Spectrum of Radiation from Inverse Compton Scattering in Blazars. The Astrophysical Journal, 561(1):111–117, Nov 2001. arXiv:astro-ph/0107152, doi:10.1086/323225.
• For the geometrical implementation of the BLR (including crossing it): Alina-C. Donea and R. J. Protheroe. Radiation fields of disk, BLR and torus in quasars and blazars: implications for gamma-ray absorption. Astroparticle Physics, 18(4):377–393, Jan 2003. arXiv:astro-ph/0202068, doi:10.1016/S0927-6505(02)00155-X.

Regarding the differences, it is a bit complicated because here we mix different implementation of the geometry, plus a different radiation field. In this plot, jetset is using a single temperature BB. We might optimize the temperature to have the best agreement w.r.t. the monochromatic line in agnpy. Or, I could implement the line in jetstet

[jsitarek]

would be interesting to really test it, but indeed it is likely that line vs full BB shape is the reason for discrepancy (in particular shift of peaks and smoothening up the sharp features in agnpy). This needs to be clearly explained in the text that the underlaying assumptions are different