1 of 19

The Anatomy of a Spectrum

S. Donnell

AAVSO Spectroscopy Section Lead

Version 1.0

Astronomical Spectroscopy

2 of 19

It Starts with the Raw 2D Spectrum

  • Light from one or more objects is spread out into its component colors across the camera sensor

​

​

​

​

​

​

​

  • ​
  • This may be a spectrum of a single object from a slit system, or a selected spectrum from many in a slitless system.

​

2

Blue End of Spectrum

Red End of Spectrum

3 of 19

The 1D Profile of the Spectrum

  • The 1D profile of the spectrum is a graph of the spectrum showing intensity as a function of pixel value

​

​

​

​

​

​

​

​

​

​

​

  • You can see the presence of spectral features such as absorption lines and noise

3

Absorption Lines

Noise

4 of 19

Sky Background and Sensor Artifacts are Included in the Spectrum

  • The raw (unprocessed) 2D spectrum includes spectral features from the sky background such as atmospheric lines

​

​

​

​

​

  • ​
  • It also includes artifacts of the camera sensor such as hot and cold pixels, some of which may be embedded within the target object spectrum

​

Sky Background Emission Lines

4

Hot Pixels

5 of 19

The Atmosphere Modifies the Spectrum Through Wavelength-Dependent Extinction

Zenith Angle = 80 deg

Zenith Angle = 60 deg

Zenith Angle = 40 deg

Zenith Angle = 0 deg

All wavelengths are reduced in intensity as the zenith angle increases

​

Shorter wavelengths (blue) are reduced more than longer wavelengths (red) for all zenith angles

​

The reduction in blue intensity relative to red increases with increasing zenith angle

​

​

5

6 of 19

Non-Optimal Orientation of the Slit Reduces Flux at the Blue and Red Ends of the Spectrum

A slit oriented perpendicular to the horizon will pass most or all of the light across the spectrum.�

A slit oriented horizontal to the horizon will have a decrease in intensity in the blue and red ends.�

Mostly affects slit spectrometers with broad spectral ranges and narrow slits.

​

​

6

7 of 19

The Spectrum is Modified by Effects of the Camera Sensor

The sensitivity of the sensor to incident light (AKA response or quantum efficiency) varies by wavelength

​

Sensors tend to be less responsive in the blue and red ends of the visible spectrum

​

Different sensor have different spectral response curves

​

​

7

8 of 19

The Spectrum is Modified by Effects of the Imaging System

Vignetting - a reduction in brightness toward the edge of the field of view compared to the center

​

Apparent in a spectrum of a uniform continuum light source

​

As a result, spectra will be fainter in the blue and red ends of the spectrum compared to the middle green part

​

Also includes variations in pixel-to-pixel sensitivity of the sensor

​

​

​

8

9 of 19

The Spectrometer Introduces Geometric Distortions

Tilt

Smile

​

​

​

9

Smile refers to the curvature of the spectral lines introduced by the spectrograph configuration.

​

A wavelength calibration frame is typically used to detect and correct for this distortion.

Tilt refers to the deviation of the spectrum and corresponding lines from the horizontal. Correcting for tilt is desired for summing values along the vertical axis

10 of 19

And, After All the Above…Noise

  • Noise is present in every spectrum you take and results from:
    • Random variations in the number of photons received by each pixel
    • Thermal electrons in the camera sensor
    • Background sky brightness
  • There are ways to reduce it, but you can never eliminate it
    • Cool the imaging camera to reduce camera sensor noise
    • Obtain images under clear dark skies to reduce sky background noise
    • Take longer exposures (up to a point)
    • Stack multiple exposures (also up to a point)

10

11 of 19

Steps to Obtain a Quality Spectrum

  1. Orient the slit to minimize flux loss
  2. Remove the sky background
  3. Remove sensor artifacts (hot/cold pixels)
  4. Reduce noise
  5. Correct for variations in pixel-to-pixel sensitivity and optical effects
  6. Correct for geometric distortions in the spectrum
  7. Calibrate the spectrum for wavelength (necessary for subsequent steps)
  8. Correct for atmospheric effects (wavelength-dependent extinction)
  9. Correct for instrument effects (sensor wavelength-dependent response)

11

12 of 19

Removing the Sky Background

Process is to define a set of regions that

    • Contains only the target object spectrum
    • Are just above and below containing only the sky

​

​

​

​

​

    • The sky-only regions are subtracted from the target object region to obtain a sky subtracted target object spectrum

12

13 of 19

Reducing Noise and Removing Sensor Artifacts

  • Remove sensor noise
    • Obtain multiple dark frames with exposure time equal to or greater than the exposure time for the target
    • Stack the dark frames to create a master dark frame
    • Obtain multiple bias frames with zero exposure time*
    • Stack the bias frames to create a master bias frame
    • Subtract the master dark and bias frames from the target frames
  • Remove sensor artifacts
    • Detect and remove hot/cold pixels with a pixel map

​

What remains is your set of target frames with the noise from the camera reduced and bad pixels disregarded

​

* Bias frames are used to scale the dark frames for various target frame exposure times.

13

14 of 19

Correct for Pixel Sensitivity Variations and Optical Path Distortions

  • Causes
    • Variation in pixel-to-pixel sensitivity in the sensor
    • Non-uniform distribution of intensity across the optical field (e.g. vignetting)
  • Correction
    • Create a set of frames of a uniform light source (not applicable for slitless spectrographs)
      • Light box
      • Sky flats
      • Dome flats
    • Create a dark subtracted master flat field frame and subtract from the target frame

14

15 of 19

Calibrate the Spectrum for Wavelength

  • A calibration frame is a spectrum of a calibration lamp or other source of spectral emission lines of known wavelengths
  • Pairing the known calibration line wavelengths with pixel value on the camera sensor they fall on establishes the relationship between pixel value and wavelength for all other wavelengths

15

16 of 19

Calibrate the Spectrum for Wavelength

  • Applying this relationship between wavelength and pixel values results in a 1D profile calibrated for wavelength

16

17 of 19

Correct for Atmosphere and Instrument Response Effects

Causes

  • Atmosphere
    • Wavelength-dependent reduction in the intensity of light (flux) due to absorption (atmospheric extinction)
      • Higher zenith angle means more atmosphere to look through
      • Atmosphere absorbs blue the most and red the least
  • Instrument
    • Non-uniform response of the sensor across the spectral range (quantum efficiency)

​

17

18 of 19

Difference Between Corrected and Uncorrected Spectrum

18

19 of 19

Steps to Obtain a Quality Spectrum

✅ Rotate the slit

✅ Remove the sky background

✅ Remove sensor artifacts (hot/cold pixels)

✅ Reduce noise

✅ Correct for variations in pixel-to-pixel sensitivity and optical effects

✅ Correct for geometric distortions in the spectrum

✅ Calibrate the spectrum for wavelength

✅ Correct for atmospheric effects

✅ Correct for instrument response

19