Landsat and the Electromagnetic Spectrum

Lesson Title Using Landsat Imagery to Investigate the Electromagnetic Spectrum: Spectral Information
Subject Area Remote Sensing Instruments – Electromagnetic Spectrum
Classification of Surface Materials
Age or Grade 8th grade
Estimated Length One class period (50 minutes)
Prerequisite knowledge/skills This presentation is intended to come near the end of the 8th grade unit on the electromagnetic spectrum. Students should already have an understanding of wavelength, the various types of EM radiation (i.e. ultraviolet, visible, infrared, microwave), and their properties.
Description of New Content Students will be formally introduced to Landsat imagery.
Goals Students will: 

  • Understand what is meant by “remote sensing”
  • Review the interactions of electromagnetic radiation and surface materials (i.e. reflection, absorption and transmission)
  • Learn about the wavelengths measured by Landsat sensors and explain why the bands are placed along the EM spectrum as they are (atmospheric absorption)
  • Understand the concept of “resolution” and how it applies to remotely sensed imagery
  • Investigate the information content of six Landsat bands (wavelength detection ranges), including blue, green, red, near-infrared and two shortwave infrared channels.
  • Define spectral signature and explain how these signatures can be used to identify and classify surface materials in remotely sensed imagery.
Materials Needed 

 1. Powerpoint Presentation
2. Know Spectral Signatures Handout
3. Unknown Spectral Signatures Handout
Procedure All imagery presented in this lesson uses Lincoln School as a focal point since all students are familiar with the area and their prior knowledge of the ground cover allows them to relate to and identify features in the images presented —This lesson could be adapted for other, more general areas if need arose.
 
Opener
Simple and complex definitions of “remote sensing” are presented. Basic interactions between electromagnetic energy and surface materials are reviewed. Students are then introduced to Landsat, an optical remote sensing satellite. Basic technical specifications are then discussed, including the wavelength range of each of the six bands and their non-coincidental relationship with atmospheric absorption, and the concept of sensor resolution, illustrated through zoomed views of a Landsat scene including the school and a comparison to a higher-resolution Quickbird image.
 
Development
Visual image interpretation. Students are shown a series of six images, representing the same ground area (including the school [built], grass and trees [vegetation] and water) in the six different Landsat bands. In the images, white values represent very high reflectance and black values represent very low reflectance, with gray tones varying continuously in between. Students are asked to compare how the different surface materials interact with different types of EM radiation and how the variations relate to wavelength. They are then introduced to band combinations, which allows information from multiple bands to be used and once and asked to again discuss the appearance of different ground cover materials.

Automated image interpretation. While much information can be gained from visual interpretation of imagery, the global coverage and relatively frequent repeat times make it impossible to produce accurate land cover classifications by hand. Instead, the information in all bands compared to signatures of known examples in order to identify materials without ever having to see the image! Students are shown the plots of reflectance across the Landsat wavelengths for two unknown materials from an area outside the boundaries of the images of their school. They are asked to guess whether the materials should be classified built, vegetated or water. After allowing them to debate for a short while, plots with the same axes for several examples from the image of the school are shown. Students are given copies of both the unknown and known spectral signature plots and again asked to now predict the class of the unknown materials.
 
Closure
The unknown materials are much more easily identifiable given examples from a known area and students are shown a high-resolution image of the location of unknown pixels to verify their predictions. In remote sensing terms, the students have just performed a “supervised classification” and “ground-truth”-ed their results. The plots of reflectance by wavelength is referred to as a spectral signature, and the spectral signatures of materials can be used to identify materials in remotely sensed imagery. Several standard examples of spectral signatures are shown, and their similarities to materials in the image of the school are discussed. At the close of the lesson, some project data relating changes in land cover class to change in population are visualized as an example of how spectral information is used in global change research.
Evaluation There is no formal evaluation for this activity. In an effort to bring complex remote sensing topics into a middle school classroom, this activity is intended to be an informal investigation of imagery and the underlying data.
Extensions Qualified students interested in participating in classification and ground-truth campaigns could be shown how to use Google Earth imagery to label examples for supervised classification efforts in the Boston area, a fairly simple but incredibly tedious activity.
References n/a

Leave a Reply