Data-driven maps take cartography a step forward: Part II Print
By Howard Veregin   
April 03, 2013

In a recent Mapping Bulletin article I highlighted some of the advantages of data-driven mapping. I argued that while vector graphics software certainly has its place in mapmaking, users should not automatically assume it is the best or only choice for making high-quality maps. Not only have map design tools within GIS improved significantly in recent years, but data-driven mapping – which relies on the underlying geospatial database to drive symbology and perform automated tasks – offers several advantages for workflow efficiency and map reuse. I argued that even greater strides would be achieved if we asked more of GIS technology, suggesting that many of limitations of current map design tools in GIS could eventually be overcome.

The article prompted several interesting email responses from readers. I received an extended comment from Ted Florence, President of Avenza Systems, the company that produces MAPublisher. MAPublisher might be described as “map production software.” Technically, it is a plug-in for Illustrator that allows users to work with GIS data in a vector graphics environment. Users of MAPublisher can import GIS data in several different formats to preserve and make use of real-world coordinates and attribute data. Since the underlying geospatial data is maintained, users can apply graphic styles and symbology based on attributes, and perform queries to support feature selection and filtering. Because of the link to feature data, MAPublisher is also able to offer automation methods usually found in GIS, including generalization and analysis tools, automated label placement, and automated legend and index generation.

Florence’s assertion is that getting a map out of GIS and into graphics software does not necessarily imply losing the underlying geospatial intelligence. Map production software like MAPublisher offers a “third way” that combines the ease-of-use and sophisticated design tools of vector graphics software with the data-driven capabilities of GIS. This includes automation tasks that can streamline the production of atlases and map series, and the ability to maintain the linkages between data and graphic styles to allow the map to be refreshed when data values change or are updated.

MAPublisher is not the only example of such map production software. Another is Ortelius, a Mac-based program from Mapdiva that relies on map attribute data for analysis and visualization. The product literature for Ortelius notes that unlike standard graphics design programs that do not recognize map attribute data, Ortelius combines the underlying intelligence of GIS with the graphics capabilities of an illustration program. Like MAPublisher, Ortelius also provides data-driven automation features, such as intelligent feature labeling.

Yet another example is Canvas from ACD Systems, a vector and raster graphics program that has an optional GIS module. This module offers tools to import GIS data, perform map projections, apply queries for feature selection, generate labels, and perform other operations based on feature properties and attributes. Like the other examples above, the GIS module of Canvas works by maintaining the linkage between map symbology and the underlying geospatial data, the key defining characteristic of data-driven mapping.

I have not used all of these map production software packages, nor am I endorsing any one of them in particular. Moreover the three examples cited above do not necessarily represent an exhaustive listing. (And I have not even attempted to list all of the GIS software with advanced mapmaking capabilities.) My purpose is to highlight the existence of map production software as an alternative to out-of-the-box GIS or vector graphics solutions for mapmaking.  Readers who want to think about how data-driven mapping might apply to their own workflows might like to consider some of these software packages as options.

The existence of map production software also lends support to the arguments I made in my earlier article. This software effectively enhances the capabilities of graphics design programs to facilitate data-driven mapping. The very existence of map production software implies that there is a market need for data-driven mapping capabilities to overcome some of the limitations of stand-alone vector graphics, especially in cases where efficiency and reuse are important. The client base for some of these software packages shows that there is a real need for data-driven mapping solutions that incorporate high-end design tools.

Indeed Florence of Avenza Systems contends that the success of MAPublisher supports the data-driven mapping thesis “precisely due to the limitations of ‘dumb’ vector graphics.” He adds that while moving GIS files to vector graphics software normally does “dumb them down,” map production software provides “a way to retain the data-driving intelligence” that permits advanced mapping and automation.

Are we witnessing the convergence of GIS and vector graphics toolsets? It’s hard to say. Predictions about the course technology will take often prove to be inaccurate. However it can be argued that the line between GIS and vector graphics software is becoming increasingly blurry. Today, many GIS packages have incorporated design tools previously found only in high-end graphics software. Meanwhile, the map production software examples described above incorporate features previously only found in GIS, allowing users to harness the many advantages of data-driven mapping.

While I’m referring mostly to traditional static paper maps, a similar line of argument might be advanced in the case of interactive Web maps, especially in light of increased interest in customizable, interactive, data-driven map displays. Earlier Web-mapping technology was based on creating a file that could be interacted with only within the limited context of the data and design elements that had been “baked in.” Today, with the advent of javascript mapping libraries and other Web technology, map data can be pulled together with symbology rules and other controls more easily to allow true user interaction and customization. Perhaps even more than traditional paper maps, Web maps embody the core idea of data-driven mapping.

At the risk of oversimplifying, I would assert that cartography is somewhat unique, in that it requires aptitude in both left- and right-brain modes of thinking. Designing an effective map entails not only creativity, expressiveness, and an appreciation for design, but also the application of logical reasoning and analytical skills to identify and extract meaningful relationships within the underlying data. It follows that mapmaking software needs to strike a balance between these two modalities, providing rich visualization and design capabilities coupled with sophisticated data analysis tools. If trends over the past few decades are any indication, we are headed in the right direction.


 
Tags: GISCartography