Remote Sensing Lab 8

Spectral Signature Analysis

Goal

This remote sensing lab provided students with skills in measurement and interpretation of spectral signatures of Earth's features.  Students collected, graphed and analyzed spectral signatures from a Landsate ETM+ image of Eau Claire, Wisconsin.

Methods

A polygon was drawn using the Drawing tool in the spectral function in Erdas Image 2013 for each of the following features.

1.) Standing Water

2.) Moving Water

3.) Vegetation

4.) Riparian Vegetation

5.) Crops

6.) Urban Grass

7.) Dry Uncultivated Soil

8.) Moist Uncultivated Soil

9.) Rock

10.) Asphalt Highway

11.) Airport Runway

12.) Concrete Surface (Parking Lot)

Once a polygon had been drawn, the Supervised tool in the raster functions was used to activate the Signature Editor tool.  The spectral signature of the image was graphed using the Display Mean Plot Window tool in the Signature Editor interface.

Results

Each of the features were added to the graph for the analysis of spectral signatures (Figure 1).

Figure 1: Spectral Signatures of 12 Earth Features

The X Axis of the graph represents the reflective bands of the image.  The Y Axis represents the wavelength of the spectral signature.  The chart below displays the highest and lowest reflective bands for the 12 features collected.

Band 1 = Blue

Band 2 = Green

Band 3 = Red

Band 4 = NIR

Band 5 = MID

Band 6 = MID

Signature	Highest	Lowest
Moving Water	Blue (1)	Mid IR (7)
Vegetation	NIR (4)	Red (3)
Riparian Vegetation	NIR (4)	Red (3)
Crops	Mid IR (5)	NIR (4)
Urban Grass	Mid IR (5)	Green (2)
Dry Soil-Uncultivated	Mid IR (5)	Red (3)
Moist Soil-Uncultivated	NIR (4)	Mid IR (7)
Rock	Mid IR (5)	NIR (4)
Asphalt Highway	Blue (1)	NIR (4)
Airport Runway	Mid IR (5)	Green (2)
Concrete Surface-Parking Lot	Red (3)	NIR (4)

Figure 2: Highest & Lowest Reflective Bands

For All 12 Features

Discussion

The graphical representation of the spectral signatures helped students to recognize that each surface feature has its own unique spectral reflectance.  This knowledge provides remote sensing analysts with the ability to identify and map features that may be unfamiliar and can also determine what bands are most useful for the identification of features through spectral signatures.  Spectral signatures can also aid in the identification and classification of images by discrete land covers.