ASTR 210 · Lecture 08

Spectroscopy and Wavelength Calibration

Turning a dispersed image into a scientific measurement

Learning Goals

  1. Explain why an arc-lamp exposure is required before a stellar spectrum can be interpreted.
  2. Compute spectral resolving power R = λ/Δλ and relate it to line separation.
  3. Identify which spectral features can and cannot be resolved at a given R.

Reading anchor: OpenStax Astronomy 2e, Chapter 5.3 (spectroscopy in astronomy).

Why This Matters

A spectrum is only a scientific measurement once its horizontal axis has been mapped from pixel position to wavelength using a known calibration source (an arc lamp with well-documented emission lines). Without that step, a spectrum is just a picture of a rainbow.

Opening Phenomenon

A raw spectral image shows a source trace crossed by a set of bright calibration lines from a companion arc-lamp exposure. Matching those lines to a reference list is what converts "pixel column 512" into "656.3 nanometers."

First question: what here is directly measured, and what is inferred from a model?

Terms for This Lecture

dispersion

Definition: The change in wavelength per detector distance along the spectral direction, often expressed in nm per pixel.

Why it matters here: It converts a line’s pixel location into wavelength and sets part of the available spectral sampling.

Relationships: A wavelength solution models \(\lambda(x)\), whose derivative \(d\lambda/dx\) is the local dispersion.

Boundary / common confusion: Dispersion is not the same as resolving power; closely sampled lines may still be broadened together by the instrument profile.

Evidence We Need to Explain

Model

Quantitative Tool

\[ R = \dfrac{\lambda}{\Delta\lambda} \]

Worked Example

  1. A spectrograph has a measured resolution element of Δλ = 0.3 nm near λ = 656.0 nm (the Hα region).
  2. Resolving power: R = 656.0 / 0.3 ≈ 2187.
  3. At R ≈ 2200, two lines separated by less than about 0.3 nm will blend; Hα (656.3 nm) and a hypothetical line at 656.5 nm would not be cleanly separated, but Hα and Hβ (486.1 nm), separated by over 170 nm, are trivially resolved.
  4. This matters directly for the next lecture: measuring a Doppler shift of order 0.5–0.6 nm requires a resolution element several times smaller than the shift itself, which R ≈ 2200 comfortably provides.

Visual Reasoning

Use the dispersion line and resolution inset together.

  • Why do the two 0.3 nm-separated lines blend while Hα and Hβ do not?
  • What would a steeper or shallower dispersion line do to the resolving power at fixed pixel resolution?
Lecture 08: Spectroscopy and Wavelength Calibrationpixel columnwavelength (nm)pixel 512 → 656.3 nmWavelength solution (arc lamp fit)Resolving power R≈2187Hαunresolved neighbor0.3 nm apart: blend into one featureHαHβ170 nm apart: trivially resolvedA linear pixel-to-wavelength fit anchors pixel 512 at 656.3 nm; at R≈2187 a 0.3 nm pair blends but Hα/Hβ (170 nm apart) never do.
Pixel-to-wavelength dispersion solution, with a resolved and an unresolved line pair compared.

Common Pitfall

Measuring a line’s pixel position and reporting a "wavelength" without ever taking or applying an arc-lamp calibration exposure from the same instrumental configuration.

Active Learning Segment

Given a short table of arc-lamp pixel positions and known wavelengths, fit a straight line by hand (two-point slope) and predict the wavelength of a science-target absorption line at a specified pixel.

Lab or Observing Connection

Lab 06 (Spectroscopic Wavelength Calibration) requires deriving a wavelength solution from arc-lamp data before measuring any target line shifts.

Synthesis

Spectral calibration is what turns a dispersed image into a scientific measurement; resolving power sets the limit on what can be told apart.

References