Discussion Questions (Week 3)

Herold et al. 2002. The use of remote sensing and landscape metrics to describe structures and changes in urban land uses.

1. What is a landscape metric?

2. What does contagion measure?
   
   

3. Describe Edge Density. How is this useful?

4. How might the fractal dimension be used?

5. Where are landscape metrics especially useful, and what are the requirements to use them?

McCauley et al. 2004. Mapping residential density patterns using multi-temporal Landsat data and decision-tree classifier.

1. What are the desired benefits from mapping land-use?

2. Why use both leaf-off and leaf-on data for the classification?

3. Why was spatial frequency information not included?

4. How were the points chosen for the decision tree training? What are the advantages and disadvantages of this method? Does this count as ground truthing?

5. Explain the likely causes of the errors in Figure 2. What are the potential implications?

6. Why might Case 5, which included all pixels within the polygon, produce less accurate results than Case 4, which used only the centroids?

7. What might you suggest as a way to improve the results?

8. How might higher resolution data affect the decision tree classifier?

Joy et al. 2002. A non-parametric, supervised classification of vegetation types on the Kaibab National Forest using decision trees.

1. What are some advantages and disadvantages to using decision trees?

2. What are the 3 selection criteria used to sample field data?

3. What is Hierarchical clustering?

4. What is a decision tree?

5. How did the authors increase the accuracy of the classification?

6. What was the study question? Was it successfully answered?