SNEMO is a series of models trained on the spectrophotometric time series of Type Ia supernovae observed by the Nearby Supernova Factory. The training procedure and results are described in Saunders, et al. 2018. Three models are presented here:

• SNEMO2: a two component model intended as means to directly compare the training data and method with other two-component SNIa models such as SALT2.
• SNEMO7: a seven component model chosen for standardizing supernova magnitudes
• SNEMO15: a comprehensive 15 component model that maximizes the amount of spectral time series behavior captured.

Usage Guidelines

While you are free to use your own fitter to use the SNEMO spectral time series templates, the SNCosmo Python library, which now includes the SNEMO models, is suggested for simplicity. If you don't already use SNCosmo, it can be installed following the instructions here. If your version of SNCosmo was installed before October 2018, you probably need to do an update.

It is recommended to run the sncosmo fit using the optional argument phase_range=(-10, 46) to restrict the fit to the phase range covered by the model. To further control the fit, the spectral time series provided for use with SNCosmo have first and last epochs added with zero flux. This helps to control the extrapolation done in SNCosmo. However, the choice that supernovae have zero flux at -20 and +60 days, which is made in the provided files, is arbitrary and is not based on the training data. The user is free to edit their copy of the templates to make different assumptions about the first and last epochs where there is measurable flux.

import sncosmo
import matplotlib.pyplot as plt

# Load some example data, or load your own data in an astropy table:
data = sncosmo.load_example_data()

# Load the model:
model = sncosmo.Model(source='snemo7') # Or choose 'snemo2' or 'snemo15'

# List the model parameters that will be fit:
fit_params = ['z', 't0', 'As', 'c0', 'c1', 'c2', 'c3', 'c4', 'c5', 'c6']

# Fit the data with the model:
result, fitted_model = sncosmo.fit_lc(data, model, fit_params, phase_range=(-10, 46))

# Alternatively, if you know the redshift of the SN, this can be fixed:
model.set(z=0.5)
fit_params = ['t0', 'As', 'c0', 'c1', 'c2', 'c3', 'c4', 'c5', 'c6']
result, fitted_model = sncosmo.fit_lc(data, model, fit_params, phase_range=(-10, 46))

# Pring the best fit values:
print "model parameters:", result.parameters

# Plot the data and the best fit model:
sncosmo.plot_lc(data, model=fitted_model, errors=result.errors)
plt.show()

Machine-Readable Tables from Saunders et al.:

Components of the three models

Model Coefficients for the training and test set supernovae (Please note: a sign error was corrected here April 24, 2019)

Tables for input into sncosmo, with lightcurve start and end-points added:

(updated April 24, 2019) SNEMO2 templates

SNEMO7 templates

SNEMO15 templates

Example spectral time series

Contact Information:

Clare Saunders (CV)
Laboratoire de Physique Nucléaire et de Hautes Énergies (LPNHE)
IN2P3/Sorbonne Université
4 Place Jussieu
75005 Paris, France

csaunders@lpnhe.in2p3.fr