Author Note: This exercise was written by Leah Meisterlin and Grga Basic (2017); revised by Leah Meisterlin (2022).
One of the most important applications of remote sensing has been tracing the transformation of land patterns and urbanization over time. The availability of Landsat satellite images (a program that has been in orbit since 1972) has made it possible for us to map land cover at different spatial and temporal scales and to quantitative analysis of land cover changes across increments since the program’s start. It is up to designer-analyst-user, however, to interpret the ground condition, to speculate, and to recognize how changes in those conditions imply the processes of urbanization, influencing a variety of interpretations and design responses. In this exercise, we will learn land cover classification methods using Landsat satellite imagery and then drill deeper into evaluating the change quantitatively.
In the early portions of the 2017 hurricane season, four major hurricanes – Harvey, Irma, Jose, Maria – ravaged the US. Since hurricane forecasting has vastly improved in the past century, many lives have been saved. However, despite the accurate forecasting, the damages incurred during the 2017 season were among the most costly in recorded history. Advances in forecasting have not stopped us from continuing to build in the floodplain.
This exercise starts with the hypothesis that in the case of Houston, Texas, the loss of wetlands (resulting from arguably rampant development and arguably lax regulations) has been a major cause of flooding that ultimately resulted in catastrophic property damage. There is a scientific consensus that wetlands can absorb large amounts of flood water. While it is fair to say that the vanished wetlands of Houston would not have prevented Hurricane Harvey’s flooding, experts agree that they would have made it a lot less painful.
Using false color satellite imagery, and referring to the Texas A&M University study referenced below as a springboard, we will visualize and measure the loss of wetlands in Houston, TX, over the thirty years between 1987 and 2017.
Jacob, J.S., et al. (2014). Houston-Area Freshwater Wetland Loss, 1992-2010. Texas Coastal Watershed Program, The Texas A&M University System.
Complete the exercise, answering the embedded questions, then produce the following four map compositions.
First, prepare three map compositions 11 x 17 inches (landscape orientation) where left half of the map represents the conditions of 1987 and the right half represents 2017. The three compositions should depict the study area portion of Houston (described in the exercise) as…
In addition, compose one map composition (again, 11 x 17 landscape) showing the wetland change from 1987 to 2017. Include in your composition the total number of wetland acres (in your study area) lost during that time. Compile your maps into a single PDF file for submission.
<aside> 💻 This section requires students to acquire the data for this exercise from the USGS EarthExplorer website. If you are not familiar with downloading Landsat data (or the other datasets offered through this website), the tutorial listed below walks through the steps of the interface.
</aside>
Some rasters have a single band, or layer, of data, while others have multiple bands. A multiband image is a collection of several monochrome images of the same scene, each of them taken with a different sensor. Each image is referred to as a band. A well known multiband (or multispectral image) is a RGB color image, consisting of a red, a green and a blue image, each of them taken with a sensor sensitive to a different wavelength. Landsat 5, for example, produces seven band images representing different wavelengths from the ultraviolet through the visible and infrared portions of the electromagnetic spectrum.
In this exercise, we will download data from Landsat 5 and Landsat 8 sensors. For your reference, the portions of the spectrum (wavelengths) included per band is summarized in the tables below.