Is Geospatial Special?

I work in a large organization. A colleague challenged me this morning regarding whether geospatial is special or not. Here's part of what I wrote. Your comments are of course welcome! Do you agree? Any other arguments or counter-arguments?

Is Geospatial Special?

Does geospatial deserves special considerations or even a special treatment within an organization?

Geospatial is a set of sciences and technologies that are often applied to specific topics, such as a geographic information system for a municipality, remote sensing for the assessment of the impacts of a flood, location-based services for a smartphone, an interactive web map to plan a trip, a virtual globe to visualize weather events, etc. Other than researchers, nobody does geospatial just for the sake of geospatial, practitioners apply geospatial knowledge and technologies to their own needs and objectives.

I'd argue that geospatial *is* special, and here's why;

  • Geospatial data, which includes at a minimum a geographic projection and datum, is not processed, analysed and disseminated by the same methods and tools as non-geospatial data
  • Geospatial data requires software that specifically supports its particularities. Software that not only must support geographic projections and datum to properly store and load the data, but it also provides the methods and algorithms for the processing and analysis of the geospatial data.
  • Geospatial data processing and analysis differs significantly from non-geospatial data. Only with geospatial data you have to consider snapping, overlapping, proximity, line of sights, position accuracy and precision, reprojections, data types like lines, polygons and voxels, specific file formats like the Shapefile and GeoTIFF, mosaicking, spatial generalization and scale change, spatial indexes in spatial RDBMS, and so on…
  • Geospatial data dissemination also differs significantly from non-geospatial data. Displaying geospatial data uses techniques specific to cartography, data is generally conveyed via maps and similar means, it requires specific data formats and web services to ensure their spatial component is conveyed along with the data, etc.
  • Trained geospatial specialists are in much better position to appropriately apply geospatial knowledge, methods and tools to solve complex challenges related to the storing, processing, analysis and dissemination of geospatial data.

In the past decade, we've seen geospatial going from being the playground of trained experts to being accessible to the general public. The advent of ubiquitous free tools such as web maps (e.g. the Google Maps API) and virtual globes (e.g. Google Earth, which has been downloaded nothing less than over 1 billion times), easy and often free access to satellite imagery, and more recently, smartphones and location-based services, educated and enabled many non-geospatially trained users to acquire and use geospatial data and technologies. This is a Good Thing ™, but does that mean that geospatial isn't special anymore? No. The reasons for considering geospatial as special stated above are still valid and geospatial is indeed special.

But does the 'geospatial' term have a future as a unifying umbrella, grouping together the domains which deals with geospatial data? Maybe not in the long run. Nowadays, people working in remote sensing, geographic information systems, location-based services and global navigation satellite systems don't necessary overlap that much anymore. Does this mean the term 'geospatial' has become obsolete, diluted into its specific traditional sub-branches? I'd say no. Geospatial is still a term useful as a way to refer to all domains which involve data with spatial references.


  1. Geospatial isn't special, but it's expected.

  2. Geospatial isn't special, it is a complex domain and most of your points could be rewritten in a similar way for other domains. I'm not going to do that here though, maybe in a full on blog post! Geospatial is special to me however, and my career is based on understanding how to build the special kind of systems you need to handle geospatial data, the way that geo based apps and services a re becoming ubiquitous is very exciting right now. Nice post, I don't usually comment.


  3. You go it right. Spatial is special.  Otherwise there is nothing special about anyone or any company doing it – and on that account, why be interested in any of them? 

    People mistake 'being preferential with being special'. 

    Jeff Thurston


  4. I think special is too loaded a word. People understand different things by it ("those kids go to the 'special' school").

    Its probably better to think of it as a domain of knowledge, which overlaps many other domains (and so is overlapped by those other domains too). Its a domain lots of people have some knowledge of, and a few people have lots of knowledge of. 

    I do note that you're arguing something is special without trying to put a firm bound on it. What is "geospatial", and perhaps more importantly, what is not? If that domain really is unique in many ways (which I'm supposing is equivalent to your argument), then a clear definition should be possible.

    On the other hand, labels aren't output though – its what you achieve with all the domains of knowledge that count. So knowledge of GeoTIFF and Shapefile formats isn't any more important than knowledge of PNG, Compound Binary File Format or whatever; unless you happen to need to achieve something with one of those particular formats. Knowledge of a spatial reference system transform isn't really important until you need to do it.


  5. Special compared to what? the word 'special' is a comparative term, so without know what A thing is being compared with it "is A special?" can't really be answered. Geospatial is special compared to flat table records like a phone books and street addresses. Not so much versus climate records or astrophysics or quantum physics or medical/biological data.

    Geospatial does hold a special place in the data continuum, serving as a coherent frame of reference that can tie many seemingly disparate data sets together and draw new inferences. Weather patterns married with addresses married with skin ailments (or rubber boot sales) for example.

  6. Just replace it with another term:

    Temporal Data are special. Events can occur before, after during. Understanding complex relationships and temporal systems (such as trying to join Earth time with Neptune time) can be very complex. Temporal columns need special software support, and can only be displayed in special ways. People must understand valid-time and transaction time which form bitemporal data. A datum is assigned to temporal data to know the units and slices of the time space.

    Ok that is enough. The fact is, storing time or geometry is complex. Using data to make sense of your world is also complex and requires knowledge of the information presented.

    What is "special" about geospatial data is not the data itself, nor the storage. It is not the data type, it is knowing what to do with it that matters, and helps classify how you use the GIS.

    For example, you may be a geographer, a statistician, a developer, etc.

    In the examples you gave above, I may not care a whit about line of site, snapping, tolerances, but just like ANY OTHER data base user I care about precision, accuracy, metadata, and data types. (Although there is WAY to much emphasis on points, lines and polygons; these are JUST ATTRIBUTES of the object you are interested in). 

    So no, its not special, its another attribute in a data base that requires modeling and good practices. What you do with data is what makes YOU special.