Interview with Dana Tomlin

In a recent conversation, Yale Professor Dana Tomlin, pioneer of Map Algebra and a founding figure in GIS, reflected on his career and how he helped shape the very foundations of modern geospatial analysis.

Dr. Dana Tomlin is the designer of the Map Analysis Package GIS, author of GIS and Cartographic Modeling, member of the GIS Hall of Fame, and originator of the Map Algebra language embodied in most of today’s raster-based GIS software, Dr. Tomlin is recognized as a leading contributor to this field.

Q: You are often called the “father of GIS.” How did your journey into GIS and geospatial analysis begin?

More like an uncle, one who has never had to assume the full responsibilities of parenthood but has always taken advantage of opportunities to play with the kids. Like several others who would also come to assume central roles in this family, my introduction was by way of coursework in environmental design.  As an undergraduate at the University of Virginia School of Architecture in the mid-1970s, I was among those who initially felt that prospects for computing in this field were just as destructive as they were constructive, and maybe even more so.  As a graduate student at the Harvard Graduate School of Design, however, I discovered that not only were those prospects compelling, but I actually enjoyed the challenge of getting an automaton to “see what I could see” on a map. I found that the design of geospatial algorithms engaged the same kind of right-brained thinking that was called for in Landscape Architecture, and a number of my algorithms that would later make their way into GIS software (for things like generating viewsheds and calculating travel costs) were initially developed in response to requests from classmates.

Q: When you first started working on GIS, did you envision it becoming such an integral tool in so many fields, from urban planning to public health?

As a matter of fact, yes. I’ve always tried to orient my work toward tools and techniques that transcend particular applications, and I’ve always found it to be especially satisfying when a procedure or perspective normally associated with one field proves to be effective in another.   

What has come as a surprise, however, is the rapid development and widespread adoption of related technologies, technologies that we now for granted but which were never foreseen (not by me at least) a half a century ago.  The Internet, smart phones, cloud computing, GPS, aerial sensors, drones: these are nothing short of magic and yet now no more exotic than what was also regarded as magic by my parents.

Q: What was the most unexpected way you’ve seen GIS technology applied over the years?

The speed and the ease with which this magic has become part of everyday life.  Initially, much of the work in this field sought to more of the real world into the lab. More recently, that work has also attempted to take lab results into the field.  And most recently, we’ve moved to eliminate that lab altogether by providing an ability to sense, to interpret, and to react on the spot without any human intervention. Among the most conspicuous examples of this, certainly, are autonomous vehicles.  And the very fact that we no longer regard them as extraordinary is something that I find extraordinary.

Q: With AI and machine learning playing a larger role in spatial analysis, how do you see these technologies shaping the future of GIS?

Regarding the impact of AI on GIS, I am at once both exhilarated and intimidated. In every sense of the following phrase, I think we are going to need all the help we can get. 

Regarding the impact of GIS on AI, however, I am very - and only - enthusiastic. Artificial intelligence is ultimately based on constructs associated with language. And GIS employs a distinctive language in which geospatial conditions and processes are represented by well- established-but-not-necessarily-optimal nouns and verbs. To my eyes, this represents an enormous opportunity. Suppose, for example, that we’d like to train a large language model with examples of a geospatial phenomenon like pollution or gentrification.  The manner in which we observe and record that phenomenon will affect the ability of any model to draw meaningful inferences about it, and this will vary considerably among our readily available options. Do we cast the world in terms of objects or conditions?  Do we focus on causes or effects?  Do we consider when and well as what and where?  And should we even concern ourselves with why?  These are the kinds of questions with which I look forward to losing some sleep.

Q: If you could give one piece of advice to students and early-career professionals in geospatial sciences, what would it be?

Just one, eh?  OK, here goes. Earlier today, I responded to an e-mail from a high school classmate whom I haven’t seen in decades. He was part of a rock band we formed back then, and he sent some ancient photos. Not only did they take me right back, but they also made me realize how much my career would be anticipated by the time I spent in that band.  The advice I would offer is therefore both straightforward and familiar:  

If your occupation is not your hobby, feel free to change one or the other.

Q: If you could use GIS for anything just for fun, what would you map?

As an academic, I am able to “use GIS just for fun” all the time, and I take every opportunity to do so. This is often in the form of homework assignments, so let me share the most recent. It reads as follows.  

Kroon Hall’s north courtyard is once again being revamped, this time by way of a generous gift from the ENV755 Class of 2025. The donors have proposed that this entire space be devoted to a distinctive sculpture, one that not only reflects well on them but also on the School - particularly as both School and sculpture respond to changing environments, activities, and perspectives. 

To help meet that challenge, your task is to indicate how this sculpture should be colored. You are welcome to use whatever colors you’d like, but their pattern must be derived from the sculpture’s form.  

It’s a maddenly open-ended assignment that becomes considerably less so when students start to tinker with that (deliberately sculpted) sculpture. As they begin to subject it to various topographic transformations, the resulting patterns begin to resemble to their own smiling faces.  The fun for me in all this lies not so much in the initial dad (or uncle) joke as in the “How did you do that?” question that follows and, much more significantly, that unmistakable twinkle in eyes of those who eventually get it.  And are beginning to get hooked.

Congratulations, Dana Tomlin, for your meaningful and successful career and your retirement! We are incredibly grateful for your passion and teachings at Yale University and wish you the best in your future endeavours.