Astronomical Spectroscopy

To generate publication images change Matplotlib backend to nbagg:

%matplotlib nbagg

Blackbody Radiation

A blackbody is a hypothetical object which is a perfect absorber and emitter of radiation over all wavelengths. The spectral flux distribution of blackbody's thermal energy depends on its temperature. Stars are often modelled as blackbodies in astronomy. Their spectrum approximates the blackbody spectrum.



In [1]:

    
%matplotlib inline



In [2]:

    
import numpy as np
import astropy.analytic_functions
import astropy.io.fits
import matplotlib.pyplot as plt



In [3]:

    
wavelens = np.linspace(100, 30000, num=1000)
temperature = np.array([5000, 4000, 3000]).reshape(3, 1)
with np.errstate(all='ignore'):
    flux_lam = astropy.analytic_functions.blackbody_lambda(
        wavelens,
        temperature
    )



In [4]:

    
for flux, temp in zip(flux_lam, temperature.ravel()):
    plt.plot(wavelens, flux, label='{} K'.format(temp))
plt.legend()
plt.xlabel('wavelength (Angstrom)')
plt.ylabel('flux')
plt.title('blackbody radiation graph')
plt.grid()

Spectral Lines

Spectral lines can be used to identify the chemical composition of stars. If a light from a star is separeted with a prism its spectrum of colours is crossed with discrete lines. This can be also visualized as flux of particural wavelengths. Flux is the total amount of energy that crosses a unit area per unit time.

There are two types of spectral lines:

emission and
absorption lines.



In [5]:

    
fig, (ax1, ax2) = plt.subplots(2, 1)

for ax in (ax1, ax2):
    ax.set_xlabel('wavelength (Angstrom)')
    ax.set_ylabel('flux')
    ax.axvline(x=6562.8, color='black', label='H-alpha', linestyle='dashed')
    ax.grid()

with astropy.io.fits.open('data/alpha-lyr-absorption.fits') as hdulist:
    data = hdulist[1].data
    ax1.plot(
        data.field('spectral'),
        data.field('flux')
    )
    ax1.set_title('absorption in spectrum of {}'.format(hdulist[1].header['OBJECT']))

with astropy.io.fits.open('data/gamma-cas-emission.fits') as hdulist:
    data = hdulist[1].data
    ax2.plot(
        data.field('spectral'),
        data.field('flux')
    )
    ax2.set_title('emission in spectrum of {}'.format(hdulist[1].header['OBJECT']))

fig.tight_layout()

Continuum Normalization



In [6]:

    
fig, (ax1, ax2) = plt.subplots(2, 1)

for ax in (ax1, ax2):
    ax.set_xlabel('wavelength (Angstrom)')
    ax.set_ylabel('flux')
    ax.grid()

with astropy.io.fits.open('data/bt-cmi-lamost.fits') as hdulist:
    header = hdulist[0].header
    start = header['CRVAL1']
    delta = header['CD1_1']
    pix = header['CRPIX1']
    waves = np.array([10 ** (start + (i - pix + 1) * delta) for i in range(header['NAXIS1'])])
    
    ax1.plot(waves, hdulist[0].data[0])
    ax1.set_title('original spectrum')
    
    ax2.plot(waves, hdulist[0].data[2])
    ax2.set_title('spectrum with normalized continuum')
    
fig.tight_layout()

LAMOST versus Ondřejov

Cross matched spectum of BT CMi.



In [7]:

    
fig, (ax1, ax2) = plt.subplots(2, 1)

for ax in (ax1, ax2):
    ax.set_xlabel('wavelength (Angstrom)')
    ax.set_ylabel('flux')
    ax.grid()
    
with astropy.io.fits.open('data/bt-cmi-ondrejov.fits') as hdulist:
    header = hdulist[1].header
    data = hdulist[1].data
    ax1.plot(data.field('spectral'), data.field('flux'))
    ax1.set_title('spectrum of {} from Ondřejov'.format(header['OBJECT']))

with astropy.io.fits.open('data/bt-cmi-lamost.fits') as hdulist:
    # waves from previous code cell
    ax2.plot(waves, hdulist[0].data[2])
    ax2.set_title('spectrum of BT CMi from LAMOST')
    
fig.tight_layout()