Lars Brodersen reviews the book Geospatial Analysis by authors Michael
J de Smith, Michael F Goodchild and Paul A Longley. "What
a book! Whether you are just a little bit interested in spatial
analysis, or you are interested on a general level, or even very
interested as specialist, this is a good book to own. For the
reasonable price of 20 Euro, this book provides a comprehensive
guidance to principles, techniques and software tools for geospatial
analysis."
Geospatial Analysis
A Comprehensive Independent Guide to Principles, Techniques & Software
Michael
J de Smith
Michael F Goodchild
Paul A Longley
Published by Matador (an imprint of Troubador Publishing Ltd)
on behalf of The Winchelsea Press
ISBN 13: 978-1906221-980 (Soft cover version)
ISBN 13: 978-1906221-522 (Hard cover version)
Web Free /20 Euro PDF / 50 Euro Print
Review by Lars Brodersen
What
a book! Whether you are just a little bit interested in spatial
analysis, or you are interested on a general level, or even very
interested as specialist, this is a good book to own. For the
reasonable price of 20 Euro, this book provides a comprehensive
guidance to principles, techniques and software tools for geospatial
analysis. Choose between a printed version (B/W), a PDF-version (full
colour, print-self) or a free web-version. The PDF-version is fully
printable. For 50 Euro the printed paperback version is available
and includes the PDF-version. Or simply check-out the free
web-version. What a bargain!
The
authors' primary concern is with developing adequate understanding of
the ways in which computer software can be used to solve geographical
problems. Therefore, the book is a comprehensive work of reference if
you need to know or want to know how to make e.g. distance
operations, grid operations and map algebra, spatial autocorrelation,
heuristic and meta-heuristic algorithms etc. The detailed
descriptions of these techniques are accompanied by plenty of
examples and tutorials. Along with the text, the book's web-site
provides a number of 'add-ons'. One example permits downloadable
datasets and spreadsheets etc. and can be used to create many of the
examples illustrated, all of which you are free to use. Another
example of the web-site's add-ons is a
set of free Powerpoint presentations available for Universities and
Colleges wishing to use these in conjunction with the book. These
Powerpoint presentations are designed to provide templates that can
be augmented with your own materials and examples, or extracted
slides can be embedded in your own Powerpoint presentations.
The
detailed descriptions of these techniques are accompanied by plenty
of examples and tutorials. Along with the text, the book's web-site
provides a number of 'add-ons'.
The book's major chapters are about 'Data
exploration and Spatial Statistics', 'Surface and Field Analysis',
'Network and Location Analysis', and 'Geocomputational Methods and
Modelling'. This book can be seen as a companion to the pioneering book
on 'Geographic Systems and Science by Longley, Goodchild, Maguire and
Rhind (Second Edition, 2005), extending or amending chapters 14 to
16 of that work. So, at first glance this work could look like a
nerd's work of reference, which it actually is. A very comprehensive
one, indeed! But it is much more than that.
Besides
being a work of reference it is a kind of general textbook on
'spatial analysis'. The first approximately 150 pages give a fine
introduction to 'Conceptual Frameworks for Spatial Analysis',
'Historical and Methodological Context', and 'Building Blocks of
Spatial Analysis'. Furthermore, the point of departure is an
introduction where different software is being discussed along with
some general reflections on GIS and spatial analysis. The best part
of this introduction, I think, is the section 'Terminology'. This
section is not a discussion about where the words
originate, but instead, provides a list of definitions of the
concepts including necessary explanations. This is something that is
so often missing in textbooks and other scientific literature, but
here is a good list of definitions of concepts.
Despite the
fact that the book is not directly a textbook you will still be able
to get a comprehensive introduction to the world of 'GIS + software +
spatial-statistics + techniques' – if you need that. If you are a
'nerd' you will be able to update your knowledge about principles and
techniques, and you will certainly have the opportunity to fine-tune
your GIS-vocabulary and spatial-analysis vocabulary. If you are a
non-nerd you will get a good chance to get closer to understand the
nerds by reading the book.
I miss a few things in the book. The book is a
guide to principles, techniques and software tools on geospatial
analysis, no doubt about that. The scope of the book is made very
clear in the foreword, in the list of contents and in the first
chapter. However, I find that a little bit more room in the book
could have been spend on the issues of input to spatial analysis and
output from spatial analysis. Chapter 3 discuss principle aspects of
input to spatial analyses, i.e. running a project based
geo-information (project's purpose, user's tasks etc.), which is
fine. My concern is that specialists in spatial analysis might assume
or seem to assume that input to their work (the spatial analysis) has
been prepared thoroughly by 'somebody else' and therefore that the
input basically is 'just there' and is kind of non-questionable. I
doubt that this in general is a sound assumption.
Therefore I
find it would have been useful if the book had laid a little more
stress on the input-aspect throughout the book. Useful in that
respect that the analysis-specialist then would be a bit more aware
of what input he starts out with, and therefore the
analysis-specialist would be better prepared to discuss input with
the project-manager, just to ensure that input in a project has been
considered thoroughly and to avoid that input is chosen by chance or
by habit. Example: The organic fertilizer (manure) from pigs is a
problem because it pollutes rivers and lakes if there is too much of
it. Therefore thorough spatial analysis must be made as a basis for
decision making to keep political control of the discharge of the
manure. One could easily think that best input to such an analysis
would be the pigs registered address (co-ordinates) and the number of
pigs at that location (multiplied by volume of manure) – a logical
conclusion.
However, if you
are not aware that the farmers trade the manure and transport it from
one end of the country to the other in order to get rid of it, severe
mistakes can be made. The manure is not necessarily put onto the soil
where the pigs are registered, often times far from it. Therefore, if
the geospatial-analysis-specialist is not considering input aspects
thoroughly, the spatial analysis can sometimes lead to incorrect
results, and waste time. The politicians will not ask what input
has been used; they will just accept the result. It is not necessary
to end in such an unhappy situation. It just needs a little bit more
awareness regarding the question of where the input actually comes
from.
Conclusions:
Get it! It is an excellent and very comprehensive guide and
introduction to spatial analysis, and there is nothing like this
anywhere else.
The other
thing, for which I miss some discussion in the book, is the output
from spatial analysis. The output of any spatial analysis must
somehow be presented in a readable form (e.g. as graphics, maps or
something else). I mean, if the output cannot be read, it is not
worth carrying out the analysis. And the more readable the result is,
the better the user will understand the contents of the result. The
problem of presenting the result of spatial analysis is not part of
the book's scope, but still, it would have been useful if there had
been at least some referencing to adequate literature (e.g. Slocum et
al's fantastic book 'Thematic Cartography and Geographic
Visualization' (new edition soon available). A
chapter or two on the output-aspect could save the reader uncertainty
about the relations between input-output and geospatial analysis.
But, it is definitely not something that reduces the qualities of this
book.
Another minor
suggestion I have for future editions is a discussion on the
principle aspects of space. The book's point of departure is, in this
respect, that phenomena can and must be positioned on the earth's
surface and referenced against Greenwich and Equator. I agree that
this reflects the general concept of modern geo-information and GIS.
But I miss a discussion what to do with non-geographical spaces like
e.g. 'an organization' and e.g. 'space of comprehension'. How does
one carry out spatial analysis on an organization, which in fact I
think is a sensible thing to do.
An organization contains
relations of many orders in the same principle manner as e.g. a road
network. The other space-aspect about 'mental spaces of
comprehension' is more complex but could be mentioned anyway. A
discussion of these aspects of concepts of space could sustain an
extension of the geoinformation-domain towards some more theoretical
ideas and concepts for the purpose of not limiting ourselves to x-y-z
coordinates referring to Greenwich and Equator.
The book is written in easy-to-understand English. The
technical aspects of readability could have been improved if
hyphenation had been used and there had been a little more
line-spacing. Relatively often the space between words is larger than
the space between lines. This means that you must be prepared to use
your index-finger when reading the text to keep track of where you
are. However, that is just a minor thing.
Conclusions:
Get it! It is an excellent and very comprehensive guide and
introduction to spatial analysis, and there is nothing like this
anywhere else.
Information: Geospatial Analysis
Lars Brodersen is Assistant Professor in the Department of Development and Planning at Aalborg University, Denmark. He is an expert in the field of Geo-communication for Geo-information. Website
Lars Brodersen reviews the book Geospatial Analysis by authors