Example 1: Predict abundance from spectra

About

This example will show you how to predict a lithium abundance from an observed spectrum by using the Spectra class in breidablik.interpolate.spectra.

Example

Lets say we have an observed spectrum of the sun and we want to determine the lithium abundance from this spectrum using the 670.9 nm lithium line.

Download the example spectrum, example_spec.txt, from https://zenodo.org/records/10428804. This example spectrum was generated from our simulation of the sun with noise added to the spectrum, how this is done is shown in example 3: generate mock observed spectrum

First, we will read in the spectrum.

# read spectra
import numpy as np
wl, flux, flux_err = np.loadtxt('example_spec.txt', unpack = True)

Since we’ve observed the Sun, we will use the stellar parameters for the Sun.

# stellar parameters for the Sun
t_eff = 5777 # K
log_g = 4.4 # cgs
met = 0 # dex

Now we can see what lithium abundance the model predicts the Sun has - for this example spectrum, 1.1 dex ⁷Li was used.

# determine the lithium abundance
from breidablik.interpolate.spectra import Spectra
# initialise models
models = Spectra()
# find the abundance and error of the observed spectrum
abund, errs = models.find_abund(wl, flux, flux_err, t_eff, log_g, met)
print(abund, errs)

Checking the value of abund, we have that the predicted abundance is 1.08 dex with a negative error of -0.03 dex and positive error of 0.02 dex, which is not a bad prediction considering that 1.1 dex was simulated.