GEOG 865
Cloud and Server GIS

Publishing original data via Fusion Tables


The Fusion Tables platform allows for the publishing of original data primarily by means of creating a new table. It is possible to upload data from an external file, but for this exercise we will create a new table in the Fusion Tables interface. In the context of this course, Cloud and Server GIS, our intention is to utilize Google Fusion Tables for its geographic capabilities, which you observed in the previous section. However, since the service is centered around a tabular data structure, that is the format in which we will create our data. This may be different from and less convenient than many other GIS data creation procedures we are accustomed to (e.g., creating and editing spatial features in desktop software, like ArcMap). However this illustrates a trade-off for relying on a cloud-based service: we are constrained by the functionality and interface designed by the service provider. Let's go ahead and create a new table.

Table 1.1
CountyName Population LatLon
Bristol 50648 41.70 -71.274
Kent 167090 41.67 -71.59
Newport 85433 41.51 -71.267
Providence 621602 41.80 -71.58
Washington 123546 41.48 -71.63
screen capture of Google Drive
  1. To begin creating a new table, go to your Google Drive home page and click New > More > Google Fusion Tables.
  2. Select "Create empty table".
  3. Click Help > Back to Classic look.
  4. Click the Edit menu and select "Modify Table Info".
    1. Change the Name to "Rhode Island Population"
    2. Click the Save button.
  5. Click the Edit menu and select "Modify Columns." You get four columns to start with (Text, Number, Location, and Date) and you are going to change these to use your own column names and data types.
    1. Change the Column Name and Type of the first three default columns to the following:
      1. "CountyName", Text
      2. "Population", Number
      3. "LatLon", Location
    2. Delete the Date field by clicking the small X next to it.
    3. Click the Save button.
  6. Populate the table with the information above by selecting "Add row" under the Edit menu and entering the corresponding values (the latitude and longitude values in the LatLon column are separated by a single space)
  7. Since we specified the LatLon column as a Location type, the values have been geocoded for us, and you should see the globe icon next to each value when you position your mouse.
  8. Click the Visualize menu and select "Map."
  9. You should see your five records represented by points at the locations specified by the LatLon values. You may need to zoom in on Rhode Island first.
  10. Clicking a point displays the attribute information from the table.
  11. Use the "Configure info window" link to customize how the information is displayed.

    You may be thinking that point symbols are not the best way to symbolize areal features, like our counties, and that polygon fills would be better. To accomplish this we need to provide the geometries of each county's boundary. We can do this by entering KML-formatted geometry information in a new column in our table. Fortunately for us, a public table with county geometries already exists, which we can join to ours, saving us the time of entering it manually.
  12. The RI County Boundaries table can be accessed via the following URL. Please open it in a new tab or window in your browser.
    1. Explore the geometry column and note that the contents are formatted as KML, and when clicking the globe icon, the outline (polygon) of the corresponding county is highlighted rather than its centroid point.
  13. To join this table to your population table, click the Merge menu and enter Rhode Island Population in the "Merge With" box.
    1. In the Merge With box, enter the table ID 1244150 and click Get.
    2. Ensure that the County Name columns are selected for both tables, indicating that they are the common attributes on which to base the join.
    3. Check the "Select subset of columns" below the county boundaries table and confirm that the Geometry column is checked. Uncheck the other columns.
    4. Check the "Select subset of columns" below your Rhode Island Population table and check all columns. This specifies that the new merged table will contain all of your original columns with the addition of the geometry information, but not the extraneous columns from the boundaries table.
    5. Check the box to "Save as a new table named" and enter the name, "RI Population Geometry" in the box.
    6. Click the "Merge tables" button.
  14. Your new table should appear and contain the following columns:
    1. County Name
    2. Geometry
    3. Population
    4. LatLon
  15. Click on the "Visualize -> Map" to display your data on a map. Notice that there are now two options in the Location drop-down menu: Geometry and LatLon. Try alternating between the two. (Remember to switch to the Classic look if you aren't seeing the Visualize menu.)
  16. To apply a meaningful symbolization to your new polygons, click the Configure styles menu.
    1. Several styles are available, but to create a simple choropleth map, click the "Fill color" option below "Polygons".
    2. Choose the "Gradient" link, click "Show a gradient" and select the "Population" column as the data to be classified.
    3. Change the From and To values to 50000 and 622000 respectively to set the minimum and maximum values in the data table.
    4. Click Save and observe the updated map symbology.

During this exercise, you have created original data using the Google Fusion tables interface, combined it with other data existing in the cloud and visualized it spatially. This represents one example of a SaaS resource that enables some amount of GIS functionality in the cloud.