Kepler.gl is an interesting geospatial mapping software to conduct Digital Humanities analysis. I first used the program with the WPA Slave Narratives to analyze where Freedmen were interviewed compared to where they were enslaved before and during the Civil War. The records are a very useful resource for researchers to obtain a glimpse of what life was like for former slaves in their own words. Nevertheless, it is important to read the narratives within the context of biases that existed within the 1930s when the interviews were conducted. Please refer to my previous blogpost on the WPA Slave Narratives and Voyant for an additional discussion on the narratives.
The narratives present an opportunity to learn more about geospatial software to visualize where the interviews were conducted from 1936 to 1938.
Kepler.gl is a relatively straightforward and user-friendly software for mapping. The open-source program requires data in the form of a CSV file before you can conduct mapping geospatial analysis. Nevertheless, there is a collection of sample data that you are free to use if you do not have a suitable CSV file on hand. If you want a visually appealing sample then I recommend using Commute Patterns in England and Wales from the sample menu.
The software requires an amount of trial and error when you work with adding different layers and features to your data. For my introduction to Kepler.gl I used the WPA Slave Narratives from Alabama. The dataset contained information such as the name of the interviewee, time and date of interview, location of the interview within the United States, and subsequent information related to each interviewee.
The user should compile the necessary information such as latitude and longitude coordinates through primary sources if it is recorded or a georeferencing website such as mapcoordinates.net.
The finished data as CSV files complete with longitude and latitude can be uploaded kepeler.gl by drag and drop or select a file on the computer.
There are various steps that the user will want to follow to create a visualization.
- After you have uploaded your data select the top left layer tab and scroll down until you see “Basic” This allows you to select a different kind of map for the data such as heatmap, time, category, cluster, and point. Each different function allows you to see how your data appears within a set location. My personal favorite is heatmap because it allows you to see where data is concentrated on a map such as voting patterns and population. Subsequent features include the time which depicts how changing time such as days or years affects the data. The basic marker is a simple point that displays the points on a geographic map without any animations or visualizations. Geographic points on a two-dimensional map cannot be placed on top of each other if their coordinates point to the same location. If you continue to scroll down the Layer tab then you can select the longitude and latitude as well as alter the color, size, and outline of points.
- Next, the user should select the base map which is located at the far right of layers and is represented by three horizontal lines. Basemap allows the user to select how they want their map to appear such as light, muted light, muted night, dark, and satellite. I recommend that the user should start with light because it is easier to see all of your data points. There are various other features that the user can click on and off by clicking the eye icon such as roads, labels, water, building, and land.
- The far right of the computer screen contains the next few important features for the software. The top right in the shape of a box with a vertical line allows the user to switch to a dual map view such as if you had a heat map and wanted to see it alongside a time map or a network map of arcs. Before you utilize the double map feature you need to go back to labels and select a second map such as heatmap and select latitude, longitude, and subsequent features from a dropdown menu. This will result in both maps being on top of each other which might make the analysis and evaluation hard to decipher which brings us back to the dual map function mentioned at the beginning of step 3.
- The icon below “switch to dual map view” is a three-dimensional representation of your current map. This allows you to see a three-dimensional representation of your data. The category below the three-dimensional representation is a legend which serves as a key to find landmarks, points of interest, and other information relevant to your dataset.
- Kepeler.gl allows the user to create unique geospatial maps designed for different purposes. One of the most beneficial map settings in my opinion is heatmap because it depicts the density of your data such as population. The Time map is subsequently useful if you are working with geospatial data across a period of time such as days or several years. The map includes a cursor to move between data on a timeline and select certain sections to play and watch as the map changes over time. Select “filters” next to layers and hover over the select a field box. Make sure you scroll down and browse your choices from the dataset. The field that can be used with the time map will have a green rectangle that says time. After your data has loaded on the time map you are free to view any time within your dataset as well as select its speed.
- The category map allows you to view different aspects of your dataset simultaneously such as former slaves interviewed in a specific location from the WPA narratives. Kepeler.gl is unique software to work with but it cannot handle a large amount of data at once for each visualization. Therefore, you need to upload your data multiple times for each category such as the location of interviews and where they were enslaved. Each subsequent dataset is represented by a different color such as one is blue, two is red. After you have uploaded the data for each category you are free to filter and pair data such as interviewed in a house with “type of slaves.” Utilize filters and select a field to continue your mapping visualization with the amount of data that you have.
- The last major type of map is the network map using arcs to link two points together such as start and endpoints. First select layers and under “Basic” scroll down until you find and select “Arc.” The next step is to ensure your latitude and longitude will work correctly to depict the data that you have available. Once you have paired the SourceLatitude, Source Longitude, Target Latitude, and Target Longitude with the correct data from your CSV by scrolling down and referring to your data you are free to create a three-dimensional mapping visualization using the top right three-dimensional map icon discussed earlier.
Kepeler.gl allows you to export your finished map as a picture, map, data, or as a URL by selecting the share icon in the top right upward arrow in the left menu bar. After you select “export image” you can change your resolution from 1x to 2x as well as your screen size. I recommend exporting your image at 2x resolution and maintaining the screen size as “original screen” or the size of your computer. Alternatively, you can export an interactive map into WordPress or on a subsequent website as depicted at the conclusion of this blog post. First, you have to select “share” and “export map” then click HTML to embed a link into your website.
Mapping is the most exciting tool I have worked with so far. It is interesting to work with mapping to answer questions on geography and see how people, land, and nations have changed and interacted over time. Researchers can utilize space as a new category of analysis and inquiry because of software such as Kepeler.gl. We can ask new questions, arrive at new interpretations, and visualize data in new ways through this technology. I still prefer using the heatmaps within Kepler.gl because it shows density better than simple points on a map. My second favorite is the Time function because it allows researchers to ask new questions about how people, land, and events changed throughout seasons and years. Mapping and Kepeler.gl allow researchers to comprehend the vast amount of interviews conducted across geography and within a short amount of time. In conclusion, Kepler.gl requires patience as well as trial and error to work properly but is worth it to raise new questions and potentially offer a new interpretation of past events.
Note: In the event that the above link does not work there is a separate link here.
Kepler.gl is located here.