Chapter 11 - Distributed GIS

Geography is about space, but GIS is becoming much less limited by distance. Although the virtual world often seems like no place, everything must be somewhere. 

11.1 Introduction

Because GIS is about geography, and therefore about space, the subject of distributed GIS is obviously multifaceted, and developing very fast. This chapter can only be an introduction to how things were at the middle of the first decade of the 21st Century.

For several years I taught a George Mason University course on "GIS and the Internet" that explored:
Because the technology was growing and elaborating faster than I could keep up, I went back to using GIS to teach science (medical geography, environmental science, data structures, etc). But this chapter is a concise introduction to these exploding possibilities.

It is helpful to think of the various components of GIS as in Figure 1.16, which organizes this chapter. But we now face a situation in which computing (whatever that is) has become part of a "cloud" of people, objects, data, and devices:

COMPONENT   EXAMPLE
Person Hospital patient with a radio tag
Software Text message with instructions on delivery
Data GPS signal
Hardware Cell phone
Network Personal bluetooth

Although ESRI is probably by many measures the the "leading" GIS provider in the world, you should be aware of some of the other sources of software, including GRASS, developed in part by David Schell, featured in Box 11.1. If you feel burdened by the cost and complexity of ArcGIS, perhaps you should check out the Open Source Geospatial Foundation (OSGEO) and Quantum GIS. (What is it about beards and GIS?)

Perhaps the simplest contemporary example of distributed GIS (and here I use the term more liberally than the authors allow) is GPS: in my kayak on the island of Kaua'i I am using and creating data on the receiver which is connected to 3 or more satellites, which themselves are interconnected as well as communicating with a Colorado control station. And when I return (assuming I haven't dropped the device in the Pacific) I can integrate my track with data downloaded from the USGS, NOAA, etc. I can also connect the receiver to a laptop to do real time "digitizing," and indeed I have taken a laptop with me on a kayak trip (but shutdown, in a dry bag!).

The advantages of this mobility to environmental science should be obvious. When Google Maps or its successor becomes easier to use, it may replace commercial GIS for many users.

11.2 Distributing the data

I call this GIS in the internet because although the data are distributed, most of the operations occur locally.

You should learn the powers-of-1000 terminology (how many zeros in a petabyte?).

Metadata has a deservedly boring reputation, but if you can't find out when, or by whom, or to what accuracy GIS layer was created, you may make a regrettable decision. I have said that if a biological lab distributed pathogens the way GIS users distribute data, the world would have died of an infection long ago. Box 11.2 for example requires basic S/T/P information - which items apply to these? Note that the other 2 qualities (scale and uncertainty) are also listed here.

While you're thinking about these dimensions, assign S/T/P to the items in Figure 11.6.

This section goes on to demonstrate that without consistent metadata "geolibraries" are a mess - imagine what a book library would be like without a catalog, or your computer without a program to organize the files?

Task: If the grid squares in Figure 11.5B are 1 kilometer, what is the scale of the map as printed on the page?

11.3 The mobile user

If we could gather together for a GPS exercise, I would stress that this relatively new technology integrates The 3 here's:
HERE example
location where you actually are on Earth
coordinates (X, Y, Z)
map where your finger is pointing

I suppose that someday when we are all (?) wearing (or implanted with?) computers it may be difficult to disentangle these 3 concepts Certainly an eagle being tracked across North America does so without conscious thought. But remember that none of this is possible without consistent coordinate systems, explicitly organized data, and a GIS that can integrate all of the components. With these, the system can presumably perform anywhere in the universe (Right Ascension, Declination, Redshift).

I have seen Reg Golledge (Figure 11.11) walking the streets of a city wearing his mobile GIS.

11.4 Distributing the software: GIServices

I call this GIS on the internet because much of the computation is happening away from the user. Consider 1) what an almost infinitely small proportion of your PC's compute cycles are being used by your activities and 2) how much either useless or rarely used software you have on the computer, and you may agree that distributing the computation - especially for something as complicated and error-prone as GIS - makes a lot of sense.

Google is almost certainly the biggest example of distributed GIS. Don't forget to look at the map of our locations!

11.5 Prospects

This chapter (and the book itself) doesn't much deal with the "geography of the internet" but as that environment becomes more important, we need to have very robust models of nodes, links, and new geometries. And perhaps one of the ironies of the geography profession loosely defined, is that in the process of taming spatial data problems we have made space matter less - what do you think?