|
|
GEO 4120c/5028c Workshop 13 April 2009 |
Introduction
Vegetation mapping, Land-cover and Land-use
classification are standard applications of aerial photo
interpretation. These tasks are useful for resource inventory, habitat
prediction, land cover/land use change analysis, environmental
characterization, assessing carbon sequestration by different
vegetation types, etc.
The two steps that must be taken before one can interpret aerial photographs in terms of vegetation or land cover are 1) one must define the question to be answered (or purpose of the mapping); and 2) one must define the vegetation/land-cover/land-use scheme that is most useful and practical for answering the question or fulfilling the purpose of the mapping. Then, one defines the scale, minimum mapping unit size, and protocols for doing the mapping.
The two steps that must be taken before one can interpret aerial photographs in terms of vegetation or land cover are 1) one must define the question to be answered (or purpose of the mapping); and 2) one must define the vegetation/land-cover/land-use scheme that is most useful and practical for answering the question or fulfilling the purpose of the mapping. Then, one defines the scale, minimum mapping unit size, and protocols for doing the mapping.
Exercise
1. Select an air photo from the
collection or, if you want to conduct this exercise using digital data,
download an interesting aerial photograph from the LABINS web site. You
will be responsible for importing the aerial photograph into either
ERDAS Imagine or ArcMap for further analysis.
2. Find
the topographic quadrangle, GoogleEarth location, or Digital Raster
Graphic of the area to help determine scale.If you downloaded a DOQ
from LABINS, the scale will be defined
3. After studying the air photo, define a research or project question (purpose of mapping), and then create the LU/LC classification system that would have the optimal level of “taxonomic resolution.” Optimal level is defined as having the most detail possible for the question AND having highest level of detectability in the air photos – i.e. what is most practical. Exercise 5 in Arnold is a good example of how to do this., and pages 41-42 have a good list of issues to consider when creating a land-cover or land-use class. Write out your research or project question, the classification, and the rationale for each class to hand in (see step 3). Be sure to differentiate land covers from land uses, and justify how you know that a particular land use exists. You may not actually know all of these until you have tried to classify your air photo. This is usually an interative process.
4.Using the aerial
photos that you selected, map the land cover/land
uses/vegetation according to your classification system. Use a
transparent overlay or create polygons in ArcMap or ERDAS Imagine (you
will have to figure out how to do this). Be sure to draw lines neatly.
Label with
abbreviations and make a legend describing what each abbreviation means
in terms of your land-cover classification. If you do this digitally,
you will have to figure out how to draw and export the map to be
printed, but don't print it - hand it in as a part of a digital file.
3. Write a
2-page report that describes your research or project question, the
classification including the justification for how you know what the
land use is, and a
description of the result - i.e. how many land covers/land uses there
are, what
the different areas are for each land cover on the air photo. Include
both the the map and the aerial photo, as the
exhibit of the mapping. Finally, answer your research or project
question.
Turn in the written report to
G:\share\GIS 4120c Aerial Photo Interpretation\your_folder, and send me an e-mail
that you have done
so. Hand in the air photo and the map at the beginning of the next
class if you use the analog method.