vector

A Crash Course on Digital Mapping for the Moderately Technologically Savvy, by Nathan Braccio

screen-shot-2016-11-03-at-4-09-50-pmFor scholars in the humanities interested in making maps there is a wide range of available tools. At least half-a-dozen programs exist that allow a scholar to upload data, visualize it, analyze it, and then share it with colleagues and the public. These tools provide the enterprising scholar the ability to augment their arguments with exciting visual components or to reveal new questions or patterns that can provide strong evidence or push research in new directions.

In this blog post I will discuss some of the options available, focusing on how each tool matches with different kinds of projects and skill levels. While not an expert in GIS or mapping, I have been working on a mapping project on 17th-century New England that has plunged me into an overwhelming array of websites and software. I made the time-consuming mistake of experimenting with each new software I came across, but hopefully after reading this post others can avoid this quagmire and get to making exciting and fun maps.

Before continuing a little should be said about the different uses for maps (from my perspective as a history PhD candidate). Maps make a striking visual argument that can both stand on its own when crafted well or can complement a text or webpage. For example, while I can point out that dozens of towns in New England were destroyed during King Philip’s War, actually mapping this destruction with intensity bubbles across the region makes a powerful statement. As an analytical tool, maps allow scholars to repurpose heavily used sources in order to find new patterns or to compile relatively insignificant data from ignored sources into more useful aggregated forms. Continuing with examples I know, by plotting something as mundane as the dates of town settlement throughout New England, the chronology of English settlers breaking away from their coastal and riverine settlements becomes clear. Simply reading dates and locations would not have yielded this conclusion. Richard White has presented a particular strong argument for spatial history.

 

Getting Started

Before you actually start to use any mapping programs, you will need a few things, including something to map! You will also need to know your goal. There are three types of things you can do with mapping software: you can make cool visualizations or tell stories, you can plot and analyze vector data, or you can overlay historical maps on contemporary maps (and even plot your vector data on them or extract data from them). Vector data is usually stored in a spreadsheet, although most programs also allow you to add data points internally in a time consuming process. With your goal in mind, gather your vector data, an image of a historical map, or information/a story you want to visualize.

screen-shot-2016-11-03-at-4-08-32-pm

If you are using vector data, you probably already have a location associated with what you want to plot. If your location is expressed in lat/lon numbers you are ready. If it is a town name or street address, you will want to convert it to lat/lon coordinates unless you are using Google Maps or Carto (they can do it automatically). A simple google search will yield some websites that will convert your locational data, but they are somewhat clunky. A better method is to upload your spreadsheet into Google Sheets (Google’s version of excel) and to create a macro. While that may sound intimidating, it really only involves copy and pasting a line of code that can be found here.

Google Maps

For the digital neophyte looking to either visualize or begin an analytical project, the best place to start is Google Maps. While you may be familiar with using Google Maps to get directions or look at a street view of your house, it also has the ability to plot vector data, or become a complex map imbedded in a website. For the purposes of the only moderately technologically savvy, this is best done through “My Maps.” My Maps has built in georeferencing and is linked to Google Sheets, so it is easy to transfer vector data over with basic locational information and have it quickly plotted through a series of intuitive commands. Your data can be classified by Google Maps in a few basic ways including simple color coding. You can also upload customized images to act as icons for your data in addition to Google’s icon library. You are able to easily manipulate your data in the program, change how it is classified or displayed, and isolate specific ranges of data with a few simple clicks.

While a more skilled user may be able to use the Google Map API coupled with coding skill and a web page to do some fancy things, for the normal user Google Maps has several limitations. Without using the API along with your own website, Google Maps has limited customizability. There is limited styling, no ability to apply JavaScript (like you can with the API), and no ability to make a customizable interface. You are also limited to using one of Google’s nine base-maps. While the base-maps can look nice, they often contain contemporary information or labels that may be anachronistic.

 

Carto

Using Carto will avoid several of these problems while requiring some additional technical skill. In general Carto is similar to Google Maps My Maps. It is good for plotting and visualizing vector data and you can modify uploaded spreadsheets within the webpage. screen-shot-2016-11-03-at-4-08-16-pmWhile at times a little less user friendly, it is also enables you to use custom base-maps and to apply limited coding to change the style and interface of your maps. Like Google Maps, it is extremely easy to share on social media or embedded in your own webpage. Overlaying images, while possible, is still a difficult task in this program. Additionally, it only has a limited ability to create a fully customized interface.

 

Arc/QGIS

Perhaps the strongest tools for analysis, although not visualization, are ArcGIS and QGIS. While very similar, QGIS is open source and free, but ArcGIS is proprietary and very expensive for those without access through their institution. These programs provide powerful sandboxes for mapping and uploading data, but are relatively difficult to use. When you open these programs, you are presented with a blank canvas.

It is up to you to upload base maps and data, or plot data in the program. Because you have a blank canvas, anything you upload needs to be georeferenced. If you are uploading a spreadsheet this means lat/long geographical coordinates, although some formats can be georeferenced within the program. The spreadsheet to be uploaded needs to first be converted into a .csv format, which can be done from excel or Google Sheets using “save as.”

screen-shot-2016-11-03-at-4-07-48-pm

Additionally, all layers of data need to have a consistent CRS (Coordinate Reference System) value applied to them.

Once uploaded, you can add to your data, categorize it in several ways, and style it with great freedom. Historical maps and images can easily be uploaded to your project and stretched and overlaid wherever and however you want. These programs are not ideal for creating visually polished maps for internet distribution (frequently people augment them with photoshop or illustrator for this purpose). Despite stylistic limitations, the Arc/QGIS alone are fine for making the mind of maps that can be included in printed publications. QGIS, which I have more experience with, also allows you to make maps into a format that you can integrate into a webpage through one of its many useful plugins. If you are interested in learning how to actually use GIS, the Programming Historian has a few useful tutorials.

 

Neatline

For mapping more focused on an extremely customizable visualization, the choices are limited unless you know JavaScript. Neatline (via Omeka) is one of the only exceptions, and as a tradeoff it has some issues uploading spreadsheets of vector data. Instead, Neatline is extremely good at creating a dynamic and interactive exhibit/story, in which the data and objects are created within Neatline. Installing Neatline is relatively simple, although it is important to know that you cannot do it on the Omeka hosted Omeka.net, but only through an Omeka platform hosted elsewhere (not a free wordpress either, meaning you would have to pay for hosting). I will not go into detail on how to install Omeka on your website here, as it is well documented here and here.

Once installed Neatline has several plugins of its own that allow you to add timelines to your map, create an interactive text in which words are linked to points on the map, and for you to upload images to overlay on the map, or even a custom background. These demos really show its power. Its greatest strength is its friendly interface and plentiful documentation. Oddly enough, some of its issues emerge when doing things that one would imagine would be relatively simple for it, like trying to batch transfer items with lat/lon associated from an Omeka database (luckily there is a helpful internet community for both Neatline and Omeka).

 

Summing Up

While I will not explore it fully here, if you have some experience with Javascript and want to work on embedding an interactive map into your webpage, there are a few different places you can start. Leaflet (QGIS compatibility with QGIS Web App), Openlayers, Google Maps API, and Timemap are all worth looking into. In fact, you will probably use some combination of these programs if you are trying to make an interactive map. As you can see, this blog post just scratches the surface of the programs available. If you are interested in analyzing old maps, take a look at MapAnalyst. If you want to make an atlas of maps you have or want to overlay multiple historic maps onto a contemporary map, check out Map Scholar made by historians. The Library of Congress made Viewshare for mapping out a digital collection on a map and storymap has a self-explanatory name.

Hopefully, this post has given those of you interested in mapping a guide on where you might want to start your project. Personally, I find myself using a combination of Google Maps, QGIS, and Neatline for different aspects of my project, with the intention of eventually taking advantage of the Google Maps API and Leaflet to bring my project online. Please feel free to contact me with any questions or suggestions.

Nathan Braccio is a Ph.D candidate in the UCONN History Department. He received his B.A. and M.A. in history from American University. His research focuses on the conflux of geography and identity in 17th and 18th century New England. More information on mapping and his research can be found on his webpage nathanbraccio.com. Contact him at nathan.braccio@uconn.edu.