Part 5—Examine Spatial Relationships Between Cities and the Urban Heat Island Effect

Step 1 Select GLOBE Temperature Data based on Spatial Location

Now that the GLOBE temperature data have been selected to include only those sites that have short grass surface cover, you will compare the temperatures reported by these sites based on their proximity to large cities. To accomplish this, first select only the sites that are within the 25 km buffer boundary and then select only the sites that are outside of this boundary. These types of relationships, such as proximity (closeness) or containment within given boundaries are known as spatial relationships between datasets.
My World Icon
  1. Launch My World GIS by double-clicking its icon on your desktop or by clicking its icon in the dock (Mac) or Launch Bar (PC).
  2. Choose File > Open Project, and navigate to the file, GLOBE_Surface_Temp_Part4.m3vz, that you either saved at the end of Part 4 or to the file that you downloaded at the end of Part 4. Select it and then click Open.
  3. To make the first selection, click on the Analyze tab, then click on the Select... By Spatial Relationship... By Containment. Select Records from the Short Grass selection, Which Are At Least Partly Contained By, Records in the 25 km buffer layer. Do Not make the selection a new layer. Name the selection "INSIDE" and click OK.
    analyze > inside
  4. Make the second selection to include only the records outside of this boundary. Again, click on the Analyze tab, then click on Select... By Spatial Relationship... By Containment... Select Records from the Short Grass layer, Which Are Not Contained By, Records in the 25 km Buffer layer. Do not make the selection a new layer. Name the new selection "OUTSIDE" and click OK.
    analyze > outside
  5. Return to the map window. In the layer list where you have the two selections, choose Highlight mode "Hide Unselected". Zoom the map in to the selections and switch between the two selections: INSIDE and OUTSIDE. Observe their locations on the map. What three major U.S. Counties with reporting sites were selected INSIDE the buffer?
  6. The major counties that are included in the 25 km buffer are Oakland County in Michigan; Lucas and Cuyahoga Counties in Ohio.
  7. What other spatial questions are raised by looking at the points inside and outside the buffers on the map?
  8. The points that are outside the 25 km buffer are also further from Lake Erie. How might this affect the temperature at the schools?

Step 2 Gather and Compare the Statistical Data

Let's return to the initial question, "Do cities experience the urban heat island effect?" By carefully controlling the variables, you are only examining comparable sites. By sub-setting and selecting data, you have excluded GLOBE sites in tropical and northern European areas, defined a city as one with over 50,000 people living in it, chosen only sites with short grass as ground cover, and finally set the area of a "city" at 25 km from the center. Now you will use the analysis capability of a GIS to gather statistics to help you draw conclusions.
  1. In the layer list, click first on the selection INSIDE the buffer and then on the Show Statistics of Selection statistics button button at the top of the layer list. A table summarizing data for that selection appears. Scroll across the table and click on the SURFT column heading to highlight it. Repeat this step for the OUTSIDE selection. Move the two windows side by side to make it is easy to compare the data.
    statistics windows only
  2. How do the mean temperatures compare?
    The mean temperature was 3.4226 degrees for the sites OUTSIDE the buffer and 4.8043 for the sites INSIDE the buffer. The mean temperatures were greater than a degree and a half apart for this one-day observation.
  3. How do the minimum and maximum temperatures compare?
    When you look at the minimum and maximum there are more dramatic differences in the datasets. Additionally, urban heat island effects are most pronounced in overnight temperatures. This is due to the fact the surfaces like concrete and asphalt cool more slowly than vegetation.
  4. Do you think this trend would hold true if you increased the size of a city to those over 50,000 people and/or decreased the buffer around a city to only 5 kilometers (3.1 miles)?
    Assuming enough schools reported data, the correlation should only get stronger and more pronounced as you increase the size of the city and as you narrow the buffer to a smaller area closer to the city center.
  5. Here is what the map looks like with the analysis completed.
    map and statistics window
  6. Save your project with a new name, such as "Urban_Heat_Island_complete".
  7. Quit My World GIS.
If you had trouble completing the steps above or would like to check your work against a completed project file, then use this project file. This file has been saved with all of the steps though the end of Part 5 completed.
Urban_Heat_Island_complete.m3vz ( 3.5MB May19 10)
Right-click (PC) or control-click (Mac) the link above to download the file.

« Previous Page      Next Page »