ESAEducationHomeWeather and ClimateGlobal ChangeNatural Disasters
   
Earthquakes
IntroductionIzmit earthquake
Floods
Thessaloniki (Greece)
Hurricanes
Storm surgesHurricane Katrina
Volcanoes
IntroductionAndean Volcanic BeltEtnaNyiragongo and Nyamuragira
 
 
 
 
 
printer friendly page
Exercise 3: Land cover and land use classification
 
Objective: To produce a thematic map showing forests, lava, snow, agricultural areas, lakes and urban centres. Also, to determine the distribution and surfaces of land cover and land use classes in the Villarrica Volcano sector of the Araucanía Region.

Data: For this exercise you will be using the same data as in Exercise 1.
 
Supervised Classification of an Image
 
Use the Villarrica-RGB-4-5-3.tif image created in Exercise 1.

The classification of a satellite image involves assigning all the pixels in an image to specific land use/land cover classes. These classes need to be predefined. This is part of the Classification process. It requires knowledge of the area, the land use, and land cover.

A classified image is therefore a thematic map that shows the distribution of selected land use/land cover classes.

In order to produce a Supervised Classification, training fields need to be identified and defined for each class.

  • forest (native and exotic)
  • agricultural
  • lake
  • volcanic lava
  • city
  • snow
 
 
The pattern of some classes
The pattern of some classes
 
In LEOWorks, select the Training Fields option via Classification/Supervised/Select Training Fields.

Draw polygones to define your training fields for each land use/land cover class in the Villarrica-RGB-4-5-3.tif image.

Save the selected training fields and classify the scene using the Maximum Likelihood method. Set Threshold to zero and select all bands for the classification.

The result is presented below. Depending on the defined training fields, your result may be different!
 
 
Supervised Classification
Supervised Classification
The selected classes are shown on the image using different colours. The colours of each class can be assigned interactively choosing Legend. The Class Statistics window contains a table showing the statistics of the classes in number of pixels and as a percentage distribution.
 
 
Task
 
Describe the spatial distribution and shapes of each class of land use/land cover identified in the processed Villarrica-RGB-4-5-3.tif image.

Describe the table showing the distribution of the six classes.

Explain the errors in some classes, both intentional and non.

Identify the distribution of classes between Lake Villarrica and the Villarrica volcano cone. Name them in that order.
 
 

 


Andean Volcanic Belt
IntroductionBackground
Exercises
Exercise 1: Display of satellite imagesExercise 2: Possible disasters
Eduspace - Software
LEOWorks 4 (MacOS)LEOWorks 4 (Windows)LEOWorks 4 (Linux)
Eduspace - Download
Villarrica.zipRegion Araucanía Google Earth.kmz
 
 
 
   Copyright 2000 - 2015 © European Space Agency. All rights reserved.