README.md

pysynch

Python interface to synchrotron libraries

This provides an interface to the C synchrotron libraries used by Hardcastle et al (1998) and in subsequent work, and python wrappers for some basic functions (synchro, see below). pysynch depends on the integration routines in the GNU Scientific Library (GSL) and synchro depends on numpy and (indirectly) on astropy as well. Building these modules has been mostly tested on Linux but it reportedly also works on Macs if GSL is installed (e.g. through brew).

installation

To install pysynch clone the git repository and do

python setup.py install

(the usual options to install may be used).

Then you should be able to

import synch

Some test code (e.g. testsynch.py, test_ic.py, test_loss.py) is provided in the demos directory.

warning

pysynch is not thread-safe.

pysynch functions

The synch library itself provides the following low-level functions as interfaces to the C code.

Units are T for magnetic field, Hz for frequency, s for time, and Lorentz factor for electron energy.

• setspectrum(gamma_min,gamma_max,power_law_index): sets the basic electron spectrum parameters. Should be called before anything else: there are defaults but they are probably not what you want.

• setpow(): sets a basic power-law model. The default; only needed if you are changing models.

• setbreak(gamma_break, delta): use a broken power-law electron spectrum with a break of delta (that is the amount by which the spectrum steepens) at gamma_break.

• setage(age, agefield): use a J-P aged spectrum with age age and ageing field agefield. Note that it's your responsibility to make sure that ageb takes account of the CMB if that's what you want.

• emiss(norm, field, frequency): calculate the synchrotron emissivity J(nu). norm is the normalization of the electron energy spectrum, i.e. N_0 in N(E) = N_0 E**-p

• cmb_ic_emiss(norm, frequency, z): calculate the inverse-Compton emissivity J_ic(nu). The redshift z must be specified.

• intne(norm) and intene(norm): integrate the electron energy spectrum over the energy range.

synchro

This code provides some of the functionality of the synch code of Hardcastle et al (1998). It allows the user to set up one or more Python instances which represent real synchrotron/IC sources and carry out certain operations on them. SSC parts of the code are not yet implemented. The code depends on numpy and (indirectly) on astropy.

To use it install the package as above and then

from synchro import SynchSource

Examples of its use are in demos/testsynchro.py.

creating an instance

To set up a source to model create an instance of the SynchSource object, e.g.

ss=SynchSource(type='sphere', gmin=10, gmax=100000, z=0.1, injection=2.0,
spectrum='powerlaw', cosmology=my_cosmology, asph=10)

cosmology here is an astropy cosmology instance, or something equivalent, which provides the methods arcsec_per_kpc_proper, arcsec_to_metres and cosm.luminosity_distance. The redshift z will be used with this instance.

Source types may be 'sphere', 'cylinder' or 'ellipsoid'. Each has its own parameters for sizes. Either physical or angular sizes may be used, but not a mixture.

• sphere: use keywords rsph for the radius of the object in metres or asph for arcsec.

• cylinder: use lcyl and rcyl (metres) or alcyl and arcycl (arcsec) for the length and radius.

• ellipsoid: use major and minor (metres) or amajor and aminor (arcsec).

Spectra may be 'broken', 'powerlaw' or 'aged'. Each again has its own parameters to pass to SynchSource:

• powerlaw: None

• broken: gbreak and dpow, the Lorentz factor of the break and the change in power-law index

• aged: age and ageb, the age (in seconds) and ageing field (in T), and optionally cmbage which must evaluate to True if you wish to include inverse-Compton losses at the redshift of the source in the ageing calculation.

Set the magnetic field in the call to SynchSource with the B keyword (or set the B attribute of the instance).

Set verbose=True to get printed reports of what the various methods do as they happen.

Initialization sets the instance attribute volume (in m^3), scale (in kpc/arcsec) and fnorm (the conversion factor between flux density in W/Hz/m**2 and luminosity in W/Hz).

instance methods

• emiss(freq): calculate the synchrotron emissivity at a frequency or a list/array of frequencies (in the rest frame of the object) and return results. The B attribute of the instance must be set. emiss uses the synchnorm attribute which will be set to an arbitrary value unless normalize has been called.

• cmb_ic_emiss(freq): calculate the CMB IC emissivity at a frequency or a list/array of frequencies (in the rest frame of the object) and return results. The z attribute of the instance must be set. cmb_ic_emiss uses the synchnorm attribute which will be set to an arbitrary value unless normalize has been called.

• normalize(frequency, flux, method='METHOD'): set the synchrotron normalization by one of several possible methods using an observation of flux density flux at observer-frame frequency frequency.

  Possible normalization methods include:

  • 'fixed': a parameter bfield must be passed -- the electron normalization is then adjusted to produce the observed emission at that field strength.

  • 'equipartition': Find the equipartition field within a given range. A tuple brange of the minimum and maximum field strengths to use must be passed to the method.

  • 'minimum_energy': Find the minimum-energy field within a given range. A tuple brange of the minimum and maximum field strengths to use must be passed to the method.

  For methods 'equipartition' and 'minimum_energy' the parameter zeta is one plus the ratio of energy density in radiating to non-radiating particles (often denoted kappa). The default value is 1.0, i.e. there are no non-radiating particles. Values < 1.0 are not physically meaningful. For true minimum energy zeta should always be left at the default value of 1.0.

  normalize sets the instance attributes B, synchnorm, bfield_energy_density, electron_energy_density and total_energy_density, which are stored in SI units. total_energy_density is by definition equal to bfield_energy_density + zeta * electron_energy_density.

synchro tutorial

Suppose we want to estimate the equipartition magnetic field in the E lobe of Cygnus A. This source is at a redshift of 0.0565 and we measure a flux of 144.42 Jy at 4.525 GHz. We approximate the shape of the lobe by an ellipse with major axis 50 arcsec and minor axis 30 arcsec. We take the electron spectrum to be a power law with gamma_min=1 and gamma_max=1e5.

from synchro import SynchSource
from astropy.cosmology import FlatLambdaCDM
c=FlatLambdaCDM(H0=70, Om0=0.3)

s=SynchSource(type='ellipsoid',gmin=1, gmax=1e5, z=0.0565, injection=2.0, spectrum='powerlaw', cosmology=c, amajor=50, aminor=30)

s.normalize(4.525e9,144.42,method='equipartition',brange=(1e-10,1e-7))

print(s.B)

This gives a magnetic field strength estimate of 4.3 nT.

Now we decide that we prefer to model with a broken power-law electron spectrum where the energy index steepens from 2 to 3:

s=SynchSource(type='ellipsoid',gmin=1, gmax=1e5, gbreak=6000, dpow=1, z=0.0565, injection=2.0, spectrum='broken', cosmology=c, amajor=50, aminor=30)

s.normalize(4.525e9,144.42,method='equipartition',brange=(1e-10,1e-7))

print(s.B)

We see that this changes (increases) the magnetic field to 4.5 nT -- which is expected since the normalizing frequency is above the break we have put in.

We find the total energy density in the lobes in a similar way

print(s.total_energy_density)

to get a value of 1.6e-11 J/m**3. The total pressure in Pa is given by dividing by 3:

print(s.total_energy_density/3)

so the equipartition pressure in the lobes is 5.4e-12 Pa.

We can plot the radio spectral luminosity as a function of (source frame) frequency:

import numpy as np
import matplotlib.pyplot as plt

frequencies=np.logspace(6,11,100)
plt.plot(frequencies,s.emiss(frequencies)*s.volume)
plt.xscale('log')
plt.yscale('log')
plt.ylabel('Luminosity (W/Hz)')
plt.xlabel('Frequency (Hz)')
plt.show()

Now we decide we want to plot the predicted flux density as a function of observer-frame frequency, labelling the normalizing point:

plt.plot(frequencies,1e26*s.emiss(frequencies*(1+0.0565))*s.volume/s.fnorm)
plt.xscale('log')
plt.yscale('log')
plt.ylabel('Flux (Jy)')
plt.xlabel('Frequency (Hz)')
plt.scatter(4.525e9,144.42,color='red')
plt.show()