search

zblz / naima

Derivation of non-thermal particle distributions through MCMC spectral fitting
http://naima.readthedocs.io
BSD 3-Clause "New" or "Revised" License
45 stars 54 forks source link

problem with Inverse Compton in Klein-Nishina regime #157

Closed nuk34 closed 7 years ago

nuk34 commented 7 years ago

Hello, Thanks lot for this great tool. I am playing around with Inverse Compton emission in the Klein-Nishina regime. When plotting the spectrum I noticed a lot of wiggles at higher energies (see IC_KleinNIshina.pdf). Is there something that can be done? Thanks a lot. Here is my example code:

PL = naima.models.PowerLaw(1e+25 / units.eV, 1 * units.TeV, 1.8)
Tstar = 39400 * units.K
u_rad = 376 * units.erg / units.cm**3
spectrum_energy = numpy.logspace(-1, 2, 1000) * units.TeV
IC = naima.models.InverseCompton(PL, seed_photon_fields=[['star photons', Tstar, u_rad]])
sed = IC.sed(spectrum_energy, distance=50 * units.kpc)
plt.figure()
plt.loglog(spectrum_energy, sed)
plt.xlabel('photon energy [TeV]')
zblz commented 7 years ago

This is a side-effect of the numerical computation method used and the quasi-monochromatic emission spectrum in the deep Klein-Nishina regime. It can be fixed by increasing the number of electron distribution points sampled per energy decade with the parameter nEed of InverseCompton, even though it will make the computation slower. There is a bit more info in #142.

A mention of this should probably be added to the docs.

nuk34 commented 7 years ago

Hi @zblz , Thanks a lot. nEed = 1000 gives perfect results at still reasonable computing time. (I should have searched a bit more in the forum to find the trick...) All the best, Nukri