Chapter 10

Spectral Analysis

Seeing beyond the visible. Learn how satellites capture unique "fingerprints" of Earth's surfaces across the electromagnetic spectrum.

At a Glance

Prereqs: Chapters 01, 09 Time: 25 min read + 25 min practice Deliverable: Index interpretation note

Learning outcomes

  • Explain what a spectral signature represents.
  • Predict how vegetation, soil, and water differ across bands.
  • Compute and interpret one simple spectral index conceptually (e.g., NDVI).

Key terms

reflectance, spectral signature, band, NIR, SWIR, index

Stop & check

  1. Why does healthy vegetation look bright in NIR?

    Answer: Leaf structure strongly reflects NIR.

    Why: Cell structure causes high NIR reflectance while chlorophyll absorbs red.

    Common misconception: NIR means heat; thermal infrared is the heat region.

  2. Why is water often dark in NIR?

    Answer: Water absorbs NIR strongly.

    Why: Absorption reduces reflected energy back to the sensor.

    Common misconception: Dark pixels always mean shadow; it can be material absorption.

Try it (5 minutes)

  1. Pick one surface (water/vegetation/soil) and predict which band will separate it best.
  2. Check the simulator and write one corrected prediction.

Lab (Two Tracks)

Both tracks produce the same deliverable: a short write-up explaining one index and what it measures.

Desktop GIS Track (ArcGIS Pro / QGIS)

Compute or visualize NDVI from a multispectral image and interpret high/low values in your AOI.

Remote Sensing Track (Google Earth Engine)

Load Sentinel-2/Landsat, compute NDVI, and export a figure. Describe what the index highlights and what it confuses.

Common mistakes

  • Interpreting an index without checking season/date.
  • Comparing scenes with different atmospheric conditions without correction.
  • Assuming a single index answers every question.

Further reading: https://gistbok-ltb.ucgis.org/

The Science of Reflectance

Every object on Earth's surface reflects, absorbs, and transmits light differently depending on its physical properties. By measuring these unique spectral signatures, remote sensing can distinguish between healthy vegetation, dry soil, and urban concrete—even when they look identical to our eyes.

The "Wheel of Fortune" Analogy

How do we identify objects from space when we can't see them clearly? It's like playing Wheel of Fortune.

Imagine you have a name with 8 letters. You don't know the name, but you reveal the first letter is J and the fourth is L.

Result: Your brain instantly guesses "JULIANA" because it fits the pattern. You didn't need all the letters to know the identity.

Satellite Translation

  • The Letters: These are the Spectral Bands (Red, Green, Blue, NIR).
  • The Name: This is the Feature (Tree, Water, Concrete).
  • The Game: Satellites only "see" a few points on the spectrum. If a pixel is bright in NIR and dark in Red, the algorithm guesses "Vegetation."

The Math of Life: NDVI

How do we know if a plant is alive from space? We use the Normalized Difference Vegetation Index (NDVI). It takes advantage of a unique biological trait: plants absorb Red light (for energy) but reflect Near-Infrared (to avoid overheating).

The Formula

NDVI = (NIR - Red) / (NIR + Red)

Interpreting the Score (-1.0 to +1.0)

  • > 0.6 (High): Dense, healthy vegetation (Rainforest, crops).
  • 0.2 to 0.4 (Moderate): Shrub, grassland, or senescing (dying) crops.
  • 0.0 to 0.1 (Low): Bare soil, concrete, or dead grass.
  • Negative (< 0): Water, clouds, or snow.
⚠️ Careful with Water! Water absorbs almost all IR light, often resulting in deeply negative NDVI values. However, algae blooms (which contain chlorophyll) can push water into positive territory, confusing algorithms into thinking the ocean is a forest!

Interactive: Spectral Explorer

Select an object and slide across the spectrum to see how its reflectance changes across different bands.

🌿 Vegetation
💧 Water
🏜️ Soil
🏢 Urban
Insight: Select an object to begin.

Summary of Big Ideas

  • Spectral Signatures are the unique patterns of reflectance for different materials.
  • Bands represent specific ranges of the electromagnetic spectrum (e.g., Red, Near-Infrared).
  • NDVI is a common index used to measure vegetation health by comparing Red and NIR reflectance.
  • Atmospheric Scattering (like Rayleigh scattering) affects the clarity of satellite imagery.

Chapter 10 Checkpoint

1. Why does healthy vegetation look so bright in Near-Infrared (NIR) imagery?

Because the cellular structure of healthy leaves reflects NIR light very strongly.
Because NIR light is absorbed by chlorophyll for photosynthesis.

2. Which surface type generally acts as an "absorber," reflecting very little NIR or SWIR light?

Dry Soil
Clear Water

Glossary

Electromagnetic Spectrum:
The range of all types of EM radiation, from radio waves to gamma rays.
Spectral Band:
A specific interval of the electromagnetic spectrum used by an imaging sensor.
NDVI:
Normalized Difference Vegetation Index—a measure of plant health using Red and NIR light.

Applied Spectral Analysis: The Band Combiner

Satellites capture data in separate "channels" or bands. By choosing which band represents Red, Green, and Blue on our screens, we can reveal hidden features of the Earth (like the "Active Composite" below).

True Color False Color IR SWIR Composite
Active Composite: True Color (4,3,2)
This combination uses the visible bands. It looks like what the human eye would see from space.
Critical GIS: The Politics of the Spectrum

Spectral analysis can reveal things the naked eye cannot—like unauthorized crops, hidden military bunkers, or mass graves. While invaluable for human rights monitoring (e.g., Amnesty International using satellite data), this same "X-ray vision" raises questions about sovereignty. Does a nation have the right to hide its activities from orbiting sensors owned by foreign corporations?

← Chapter 09: RS Foundations Next: Chapter 11: Image Classification →

BoK Alignment

Topics in the UCGIS GIS&T Body of Knowledge that support this chapter.