Description

Cons 127 Observing the Earth from Space

Lab 5: Detecting Change with NDVI: Vegetation Responses through the Spectrum.

Total marks possible: 29

Instructor: Prof. Nicholas Coops

Office: FSC 2301

604 822 6452

nicholas.coops@ubc.ca

TA: Dave Williams

Office: FSC 2232

Dave.Williams@forestry.ubc.ca

For this assignment, you will...

•  Learn about the spectral responses of vegetation
•  Learn how to calculate and utilize the Normalized Difference Vegetation Index (NDVI)
•  Be introduced to analyses of land cover, land use patterns, and land cover change
•  Explore how humans are impacting terrestrial ecosystems

Submit the following in a report:

• Answers to questions 1 through 8. Brevity is much appreciated!
•  When answering these questions, you DO need to cite your references and put answers into

your own words. (2 marks)

This lab requires a computer with an internet connection, and the Java Web Application, which you used

in Lab 4 (but can be downloaded here). If Google Chrome is your default browser you’ll need to switch

to Internet Explorer if using Windows, or Safari if using a Mac OS. If you have any outstanding issues

running Java talk to the TA to set up a time to complete the lab on a Forestry computer.

Part 1: Interpreting Change using ESRI Change Matters

1. Understanding how Earth systems change through time, either due to natural or anthropogenic

processes (or both!), is an integral part of understanding how the world around us operates, how it

affects us, and how human impacts the natural world. With increasing data availability and

abundance, newer and better sensors and satellite constellations, and an increasingly long timeline

of observations, remote sensing scientists are well-positioned to detect a change on the Earth at

many resolutions – in time, space and the EM spectrum. Satellite sensors provide consistent and

repeatable measurements of everything from whole-earth phenomena, like global primary

productivity maps derived from MODIS data, to sub-meter imagery of the trees lining city streets.

2. One common way of remotely sensing change on the Earth’s surface is by analyzing vegetation

through time – its spatial patterns and abundance, as well as its spectral properties. Because

vegetation reflects NIR wavelengths strongly (remember the red edge from lecture), remote sensing

scientists have developed vegetation indices, the most well-known of which is probably the

Normalized Difference Vegetation Index (NDVI), that take advantage of vegetation’s spectral

qualities. Using spectral vegetation indices we can directly and quantitatively compare remotely

sensed images. Vegetation indices also allow us to assess plant health and vigor and can be

combined with other information to, for example, assess how surface temperatures are changing

due to the expansion of Indonesian oil palm plantations (Ramdani, Moffiet, & Hino, 2014), or study

how cities grow and change.

Q1. What is the formula for the NDVI (1)? Which Landsat bands are used to calculate the NDVI (1)?

3. In a web browser, I got to http://www.esri.com/software/landsat-imagery/viewer. You’ll initially be

shown three false-color images of Mount St. Helens: one in 1975, one in 2000, and one showing

the NDVI change between the first two images. To the top-right of the images is a “View Larger” link

that you might want to click, though note that it may cause the three images’ views to become

disconnected.

Q2. What changes do you see in vegetation between the two images (1)? Where has vegetation

increased (1)? Where has it decreased (1)? What are the possible reasons for these changes (2)?

4. Click on the Select Dates dropdown menu, and choose “1975-2010”. Next, type “Surat” in the

Search dialogue box in the top left of the screen and hit enter/return. When the “Choose a location”

a window pops up, select “Surat, Gujarat, India.” With a 4.8% average annual growth from 2010-

2015, Surat is one of the fastest-growing cities in the world (UN Economic and Social Affairs, 2014).

Zoom each image in and out by one or two clicks (if they’re not all linked together), and examine the

image features.

Q3. Based on the legend at the bottom of the window, how has land-cover changed as a result of

Surat’s expansion? (2)

5. In the Select Dates drop-down menu select 1990-2010. Next, in the Select Image Map drop-down

menu, select Vegetation Analysis. This false-color representation still highlights vegetation but

also makes other features, such as roads, more visible in the image.

Q4. In which directions are Surat expanding (1)? What clues can you see in the image that might help to

explain why Surat is expanding the way it is (2)? Justify your answer.

6. Type “Nootka Island, BC, Canada” into the search bar and hit enter/return. Under the “Select

Dates” dropdown menu, select 1975-2010. A new infrared image should load (click okay on any

warning messages that might pop up). Zoom in until the island more or less fills each image frame.