Part 4—Prepare Data for Analysis

Step 1 – Explore U.S. County Population Data in Ohio to Consider Factors That Might Affect Surface Temperature

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 (Win).
  2. Choose File > Open Project, and navigate to the file, Urban_Heat_Island_Part3.m3vz, that you either saved at the end of Part 3 or to the file that you downloaded at the end of Part 3. Select it and then click Open.
  3. When the project opens, the view is zoomed into Ohio and the following layers are in the Layer List:
    • Globe Surface T 12 03 2008
    • U.S. Cities > 50,000 (turned off)
    • U.S. Cities (turned off)
    • U.S. States
    • U.S. Counties
    • Countries
    • MODIS_landsurf12_08.tiff (turned off)
    13 amp zoomed to Ohio
  4. Turn off the Globe Surface T 12 03 2008 layer.
  5. Edit the appearance of the U.S. Counties layer to depict counties by population using a grayscale color gradient. On the U.S. Counties layer, click the drop-down menu next to Fill Color and select Population (2000). Notice that once you do this there is a legend for this layer at the bottom of the map.
    6. Click and hold on the drop-down menu next to the words Fill Color and select Population (2000).
    1 select population
  6. Turn on the Category List (legend of the layer) by clicking the L and selecting Name from the drop-down menu. The map will now also display the names of the counties.
    2 counties with legend displayed

  7. Make U.S. Counties the active layer. Then click on several counties with the Get Information Tool identify to discover names of the counties with the largest populations.
    • Which county in Ohio has the largest population?
      • Cuyahoga has the largest population. 1,393,978 million people lived in Cuyahoga county in the year 2000.
  8. Use the Move Map Pan and Get Information identify tools to find the most populated county in other states in the area. To see the name of the county displayed in the Legend to the right of the map, choose the Pointer tool on the toolbar and click on various counties.
  9. Turn on the U.S. Cities > 50,000 layer. Make it the active layer, open the table of the layer, and sort the table in ascending order by the State field. Scroll down the table to find the cities in Ohio with populations greater than 50,000.
    • Look at the map, how do the large cities compare with the most populous counties?
      • The larger cities are clustered in the more populous counties.
    • What factors of large counties do you think will lead to increases in surface temperature?
      • In counties with large populations, there are greater amounts of man-made surfaces, such as parking lots and buildings. Dense urban areas generally lack parks and trees, as well as open and vegetated space. Both of these factors can contribute to increases in surface temperature.
  10. Turn off the U.S. Cities > 50,000 layer.

Step 2 – Subset GLOBE Data to Create a New Surface Temperature Layer for Michigan, Ohio, and West Virginia

  1. Turn on and make the Globe Surface T 12 03 2008 layer active. Click the Show Table of the Active Layer button to explore the data in the layer.
    2. While exploring, consider the following questions:
    • Where is the greatest concentration of schools reporting surface temperature data?
      Ohio has the most schools reporting, but the area including Michigan, Ohio, and West Virginia contains almost all the data points.
    • How many different countries reported data for this day?
      Three countries reported: U.S., Poland, and Trinidad and Tobago.
    • Would it be valid to compare data from West Virginia to Trinidad and Tabago? Why or why not?
      No. This would not be valid because the latitude, and consequently, the length of the day and directness of the incoming solar radiation are very different for these two places.
    • What variables do we need to consider when comparing sites?
      Latitude, elevation, and land cover types are some variables to consider.

  2. When you are done exploring, close the table of the layer.

  3. In order to study the area with the most data and also to limit the variables in the data, you will narrow the focus to the schools in Ohio, Michigan, and West Virginia. Click on the Analyze tab. In Analyze mode, choose Select...By Value.
    In the panel to the right, make the following choices:
    • Select Records from: Globe Surface T 12 03 2008
    • Whose: SP (state) Matches One Of The Values Checked Below. Check the boxes next to MI, OH, and WV.
    • Check the box Make Selection a New Layer. Type "Surface Temp in MI, OH, WV" in the Result Name field. Click OK.
  4. A new layer will appear on the top of the Layer list. This new layer is a subset of the original one, it contains only the GLOBE surface temperature data for the selected states. This map shows the result of the selection.
    6 both layers on map
  5. Turn off (hide) the original GLOBE layer (Globe Surface T 12 03 2008) with all the data.

Step 3 – Select Short Grass Surface Cover Records From the MI, OH, WV Surface Temperature Layer

surface cover types
GLOBE surface temperature data can be submitted for a variety of surface cover types including short and tall grass, barren land, shrubs, dwarf shrubs, concrete, asphalt, open water, and other.

Comparing the surface temperatures on asphalt to grass would be like comparing the temperature of black versus white cars on a sunny day. Therefore, to make a scientific comparison it is necessary to limit the comparison to only one surface type that is common at schools: short grass. To do this, you will further subset the data, limiting it to only short grass cover.

  1. Click on the Analyze tab. In Analyze mode, choose Select...By Value.
    In the panel to the right, make the following choices:
    • Select Records from: Surf Temp in MI, OH, WV
    • Whose: SCOV (surface cover) Matches One Of The Values Checked Below. Check the box next to "1"
    • Do NOT check the box Make Selection a New Layer. Name the result "Short Grass". Click OK.
  2. The GLOBE records that are located in one of these three states and that also have a cover type of short grass have now been selected.
    short grass selection results
Short grass was the cover type selected here, but this same study could be replicated using other commonly reported surfaces such as concrete or asphalt.

Step 4 – Create a 25-Kilometer Buffer Around U.S. Cities With a Population Greater Than 50,000

Imagine you are driving out of a large city and into the countryside. How far from the center of the city do you have to go until the landscape begins to change from urban to rural? Is this answer the same for every city? What data or imagery could help you identify that boundary? Aerial or satellite imagery could provide answers on the use of land. Both NASA and the USGS maintain the LANDSAT satellites that measure land use. Since there is no perfect answer for this distance, you will start by creating a 25-kilometer boundary, or buffer, around each of the selected cities with populations over 50,000.
  1. To make a buffer around a set of points, first click on the Analyze tab, scroll down the list of options, and then click on Make Buffer Around....
    In the panel to the right, make the following choices:
    • Create Buffer of: U.S. Cities > 50,000
    • Set the buffer at 25 Kilometers (about 15.5 miles)
    • Click the radio button for Dissolve All Buffers
    • Name the result "25 km of Cities". Click OK.
  2. A new layer is created for the 25 km buffer, and displayed on the map as 25 km of Cities. Edit the appearance of the new layer to make the buffers transparent. Select the new 25 km layer by clicking on it, and then click on the Edit Appearance edit layer appearance button button just above the Layer List. In the window that opens, select the following options:
    • Set Choose Fill Color by to Uniform and make it a pale orange
    • Adjust the slider for Transparency to 70%
    • Change the Outline Color to Orange
    • Adjust Outline Width to a larger option
    • Click Apply and Close.
    1. This window shows the options for changing the Fill Color when editing the appearance of a layer.
      10.1 edited appearance of buffer


    2. This window shows the options for changing the Outline Width when editing the appearance of a layer.
      10.2 edited buffer



  3. Visually compare the pattern of GLOBE surface temperature data with the 25 km buffer around U.S. Cities with a population over 50,000.
    Part 4 final map

    4. The urban heat island effect is theorized to affect large cities that have converted a significant percentage of the natural vegetative cover to roads, parking lots, or buildings. Do you think there is a specific population level that is the definitive tipping point between cities that are affected by urban heat islands and those that are not? What population value would you choose? In what ways could even large cities combat excessive heat build-up through their planning or projects?

    The EPA uses a population of 1 million in their examples for large cities and 50,000 was used as a starting point here. However, this study could be completed using various population levels. Urban heat island severity depends greatly on the amount of natural space preserved in an area, the use of green or living roofs, the use of concrete instead of asphalt for roads, as well as other factors which reduce the amount of absorbed energy.

  4. Click on Surface Temp in MI, OH, MW to make it active. In the layer name, under Highlight Mode, click the All (highlighting off) radio button. Then use the Pointer tool and click on the points on the map to investigate the temperature records inside and outside of the buffers. How do these temperatures compare?
    There appear to be more warm temperature records inside the buffered zones.
  5. Save your project with a new name, such as "Urban_Heat_Island_Part_4".
  6. Quit My World GIS.
If you had trouble with any of the steps of this part, such as sub-setting the GLOBE surface temperature data, limiting the cover to short grass, or creating the 25 kilometer buffer around the cities or if you simply need a fresh project file with all the data for the next part of the chapter, then use this project file. This file has been saved with all of the steps though the end of Part 4 completed.
Urban_Heat_Island_Part4.m3vz ( 3.5MB Jun2 10)
Right-click (PC) or control-click (Mac) the link above to download the file.

« Previous Page      Next Page »