In the book Geospatial Analysis, authors Michael J. Smith, Michael F. Goodchild and Paul A. Longley suggest “geospatial analysis” exists in conceptual terms as “a subset of techniques that are applicable when, as a minimum, data can be referenced on a two-dimensional frame and relate to terrestrial activities.” They reason that the frame is the integral component here, and that “everything is nowhere” without it. They go on to explain that statistics (descriptive, exploratory, and explanatory statistics) should be considered and that strict 2D analysis alone remains too restrictive. Thus, geospatial analysis includes a 3D consideration. Additionally, visualisation is also considered integral to geospatial analysis.

 If we take 10 geospatial professionals and ask them to define geospatial analysis, I would be willing to bet 10 Euro that we would get at least 5 different answers. It is difficult for many to define.

Nevertheless, I think we would find many people trying to explain products or services that can be broadly labelled as ‘descriptive, exploratory or explanatory’ in nature. These terms would encompass surface modeling just as well as they could include Virtual Earth web maps, GIS buffering, CAD cross section layouts, places Inuit live, land use patterns or the route an oil tanker is taking through the North Sea.

Even a quick cursory glance through the Index of Geospatial Analysis will encounter terms like measurement scales, geometric means, resolution, stream networks, variogram, breaklines, ISODATA and percentile, for example. Yet, you will not find the words Google Earth, Virtual Earth or virtual reality (VR) in that index. Is geospatial analysis hard science alone? Is it solely in the domain of geo-computation, mathematics and statistics?

We need to ask ourselves, what is the role of traditional science such as geo-computation, mathematics and statistics in modern day geospatial analysis and what is the role of some of the new technologies (often generated from traditional geospatial analysis) like virtual worlds, VR and probably robotics and artificial intelligence (AI) to come soon.

When I think about geospatial analysis I see it in terms of a constellation of products like GIS, CAD, hydrological modeling software, lidar point cloud analysis software and even GPS base station software and processing and so on. Hardware, software, media and approach all contribute to description, exploration and explanation. 

Ultimately, we are faced with the question of quality and re-use when talking about geospatial analysis. And when we start to talk about quality and re-use, then we begin to see that some data cannot be used for geospatial analysis because it is not very good data; either we don’t know where it came from, how it was collected, when it was collected or even if it is in the database correctly etc. 

It is important to recognise that a huge amount of geospatial analysis is going on, particularly due to the positive impacts of virtual worlds, VR, gaming and robotics etc., and that each of these is raising geospatial awareness around the world immensely - importantly raising spatial literacy in the public mind, and eye.

But, we need to nudge our neurons to understand (and explain to others) the distinction between data types, data quality and reuse factors and how these correlate to mathematics, statistics and so on, and that value in a wider landscape of describing our world, exploring it and explaining all the things going on in it – the least of which is a need for high quality data analysis to solve real problems.

We must embrace the fact that a real need for quality data and analysis is a critical necessity in resolving basic physical and environmental processes, building critical infrastructure and securing our cities, for example. And that this implies high quality spatial analysis. And we need to leverage the spatial awareness gained through virtual worlds, VR, games and so on, with the realities of terrestrial processes. This is not an either / or proposition.

Information:

Geospatial Analysis: Web site, PDF and Book

This week’s question contains a broad statement that needs clarifying. I’m mainly addressing the traditional geospatial crowd where considerable time has been spent over the course of many years to stand up a system to meet the specific needs of an organization. The purpose of the question is to draw out GIS practitioners to discuss the level of engagement with their collected data.

Analysis certainly takes place with all GIS systems, and with most GIS practitioners. However, is there enough spatial analysis going on? Is the geospatial data being put to good use to provide decision support and business intelligence? Is spatial analysis adequately addressed by today’s practitioners to make the most of the time and effort that goes into data creation and data maintenance?

GIS Analysts

My sense is that a majority of the time spent by GIS analysts goes into data creation and maintenance. Administering a spatial database while maintaining data quality throughout an organization is not a trivial task. The demands of system maintenance and database expertise is considerable, requiring a constant update of knowledge regarding hardware and software requirements to achieve optimum performance.

There’s also the issue of dealing with many different data types and working to assure quality and accuracy while adhering to established standards. Ingesting data from disparate sources, in both raster and vector format, and converting and manipulating it is a time consuming task.

The integration of GIS with other systems to provide map-based views of information is another important pursuit that extends the reach of GIS within an organization. This often involves the publication of data to the Web and the creation of customized applications. This skill set often requires strong programming knowledge and the ability to stand up customized systems that match the needs of internal clients. Often times project management is necessary with outside vendors and internal IT partners. The resulting systems make good use of the data that has been collected, and add new data that is now correlated with place.

There’s also the extension of GIS data into the field for real-time access and update of information. The field-based use of data with extension to the fieldworker provides a great deal of efficiency and adds considerable value and accuracy to the data.

Increasingly GIS practitioners are asked to model their data in 3D, and to extend their data into 3D platforms (such as with KML). The increased realism of these environments adds considerable awareness and intuitive interaction, adding greater context and improving understanding.

GIS analysts are often tasked with the creation of map products. These static views of information are important communication tools. As an output of GIS, these products can be regularly updated to provide needed information regarding specific metrics.

The above geospatial tasks, which are often conducted by multiple individuals within an organization, involve very little spatial analysis. In a sense, the creation of thematic map products for distinct purposes is a means of analysis as it performs a function with geospatial data that aims to answer questions, but what’s missing is the in-depth spatial modeling, spatial statistics and spatial data mining that prove the power of GIS.

Analysis Tools Provide Answers from Amalgamations

I’ve gained an appreciation for map analysis over the years by working with Joe Berry, and editing his latest book, “Map Analysis.” Joe is the king of communicating ideas on ‘map-ematical’ processing to make this reasoning, operations and analysis accessible to all practitioners.

The mathematical world of spatial analysis can be daunting to most practitioners, particularly as presented in academic journals, with theory isolated from practice. That’s why Joe’s work is so popular and important, because it ties theories to examples and walks the user through the reasoning.

The concept of industry-specific data models touted by ESRI and others goes a long way in establishing a solid foundation for industry-specific modeling. ESRI’s geoprocessing framework, with exchangeable data processing scripts and models, speeds the creation of knowledge from information. And Spatial Analyst is a mainstream tool set that provide a good deal of insight into data. While the spatial analysis toolset has improved over the years, there’s a considerable way to go to make in-depth spatial analysis more accessible and commonplace.

Analysis tools will really begin to shine when they’re coupled with a collaborative system of system view that combines the knowledge and data from multiple disciplines for a holistic understanding of system interactions.

Additionally, there are whole frontiers of spatial analysis that have yet to be explored. The addition of time for 4-D GIS adds whole new analysis possibilities with temporal analysis. And the move to 3D means there will be a whole new area of exploration and insight that unveils insight based on proximity in 3D space.

Never Bored?

I don’t mean to disparage the current state of practice or belittle those that don’t have time or institutional inclination to conduct in-depth geospatial analysis. But I do want to point out the vastly unexplored possibilities that make future practice so exciting.

A comment that I read some time ago about journalists has stuck with me, “we may not always be happy, but we’re never bored.” When writing, there’s always a different angle or a new source to uncover. The limitless possibilities of GIS work strikes me as a similar parallel.

There are certainly an endless number of customized systems and portals to be built, more stakeholders to engage to show the merits and possibilities with the software. There’s also a growing need to work with data from other disciplines in order to conduct analysis that gives us greater insight into how we use and manage the land.

A great deal of time and effort have gone into establishing considerable geospatial data stores with increasing levels of accuracy. We’ve only just begun the exploitation of this stored data in order to solve real-world problems. The era of spatial analysis is in its infancy.