spatial analysis and spatial statistics for research in geography, gis and earth management special...
TRANSCRIPT
Spatial Analysis and Spatial Statistics for Research in
Geography, GIS and Earth Management
Special Lecture Series
Henan UniversityMay and September, 2012
Dr. Ron Briggs (call me Dr. Ron)Professor Emeritus
The University of Texas at Dallas 1
Introducing the Lecture SeriesToday, I will introduce the lecture series: Introduce myself Explain the goals of the lecture series Discuss fundamental research principles:
– the major steps of any research project
– the importance of the research objective
– the need to review the existing literature Introduce a project that you can work on
2
Geography, GIS and Earth Management
Geography– The science of location– Where things are located, and why
Geographic Information Science (GIS)– Technologies for analysis of where and why
Earth Management– Geography and GIS applied to the real world
3
Why
?Where
This class is for you all!
4
Description Explanation Application
Geography
GIS
Earth Management
Geography, GIS and Earth Management(G-GIS-EM)
This class is for you all!
Who am I?
Dr. Ronald Briggs (call me Dr. Ron) Professor Emeritus (retired!?)
– University of Texas at Dallas Email: [email protected] or [email protected] Web: www.utdallas.edu/~briggs Office: Room 319
– Come and talk with me there– My cell phone number is 18739962744
5
Personal (continued) Born in Liverpool, England
– At the same time as The Beatles
6
Moved to the US in 1966– Ph.D. The Ohio State University, Columbus– First job at University of Texas at Austin– Moved to UT-Dallas in 1976– Director of computing (1982-1995)– Director of programs in GIS (1995-2008)
AustinDallas
Columbus, Ohio
Texas
7
My specialization is Urban Geography and GIS
Rio de Janeiro, Brazil
Nanjing Rd, Shanghai
Sydney, Australia
8
But bustNorth Island, New Zealand
But I am also interested in the natural environment
Amazon Rain Forest, Brazil
British Virgin Islands, Caribbean
9
The Goals of the Lecture Series
10
Goal #1 help you conduct high quality research in
Geography and Geographic Information Systems
describe standard research practices and procedures in western countries
11
Goal #2: Improve your English.
By talking in English about some GIS concepts you already know
and some that you don’t know hopefully, I can provide some new insight
and understanding!
12
Goal #3: Learn about Spatial Statistics
We will focus on Spatial Analysis in general, and Spatial Statistics in particular.
Hopefully, you already know at least something about spatial statistics
Talking about it again here will help your English, and help you better understand spatial analysis and spatial statistics
n
)x(x
n
)x(x
w/)x)(xx(xw
n
1i
2i
n
1i
2i
n
1i
n
1i
n
1jij
n
1jjiij
Spatial Autocorrelation
Centrographic Statistics
13
Goal #4: Learn about doing research on the U.S.
Learn about data sets for the United States Learn about geographical concepts used
with US data, for example– Coordinate systems such as the State Plane
Coordinate System– Geographic units used for reporting data such
as Metropolitan Statistical Areas (MSA), census tracks, census blocks, etc.
14
Research in geography, GIS and Earth Management
Types of Research Spatial analysis—the focus of my lectures The steps of a research project The importance of the research objective Using and citing the Existing Literature
15
Types of research in G-GIS-EM
1. Data set generation• Often what you do for your first job!
2. Technology or tool development• Usually at universities or computer companies
3. Spatial Analysis• identifying spatial patterns
• understanding why they occur
16
?
but all are interrelated
Research Interrelationships
17
DescriptionExplanation App
licati
on
Technology or tool development
requ
ires
used
for
requires
Spatial Analysis
uses
used for
used forData generation
Example: Urban tree inventoryFang Qiu, Jie Chang , Caiyun Zhang at UT-Dallas
Urban forests (trees in cities) very important resource today
Need inventory of trees by tree species– Data generation
Interpolation and cluster analysis techniques used to identify trees and their species– Tool development and spatial analysis
18
Interpolation and cluster analysis techniques applied to remote sensing data used to identify tree species
Ground Points
Accuracy doubled from existing methods: --60%-70% versus 30%-40%
Urban tree inventory:relevant to all your areas of study
Physical geography– Trees are an important part of the natural environment– Biogeography and forestry
Economic geography– Trees are an important economic resource – significantly increase the value of houses and land
GIS– GIS tools and techniques required to do the inventory
Earth Management– Trees are a critical resource to be managed
21
$
Briggs Henan University 2010
Location of children with cancer --they appear to cluster(concentrated in one area)Why?--because people cluster in cities?--because there is a source of pollution?
Example: cancer data(source: Rogerson and Yamada p. 17, from Cuzick and Edwards, 1990)
(data is for childhood cancer in northern England)
Need to apply spatial analysis to examine if cancer cases still cluster after we remove the obvious factor that people cluster in cities.
The focus of the my lectures
23
This lecture series will focus on tools and techniques for spatial analysis
The goal of spatial analysis is to− Describe and identify spatial patterns, and − understand why they occur Spatial analysis usually
– states a hypothesis: a guess at a possible explanation e.g. Environmental pollution results in higher rates of disease
– designs a study to test that hypothesis using Appropriate data Appropriate tools and techniques
Spatial statistics are often the best tools to use– We will focus on spatial statistics– See Outline of Lecture Topics
•
Stating an hypotheses– A hypotheses is “a guess at an explanation,” for example– (More people get sick in areas which have more pollution)
Designing a study to test that hypotheses
24
Outline of Lecture Topics: Spring (May)Week 1 (May 7)
1. Research Methodology: The Central Role of Goals and Objectives. (1Intro.ppt)
2. Spatial Analysis: Concepts and Issues (2spatanal.pp)
3. Spatial Data: What is special about spatial data? (3spatdata.pp)
Week 2 (May 14)
4. Standard Statistics & Spatial Statistics: Differences & Similarities (5CentroStat.ppt)
5. Descriptive Spatial Statistics for Points and Polygons
6. Inferential Spatial Statistics: Standard and spatial (6InfStat.pp)
Week 3 (May 21)
7. Point Pattern Analysis: Concepts and Tests for Clustering and Dispersion (7.1PointPat.pp)
8. Point Pattern Analysis: Applied Point Pattern Analysis (7.2PointPat.ppt)
9. Conclusion and preview of September classes: Analyzing polygons and surfaces
25
Outline of Lecture Topics (Fall)Week 4 (Fall- 1st week)
10. Review of May sessions (Point Patterns) and Intro to fall (Polygons & Surfaces)
11. Spatial Autocorrelation: concept and implementation (9SAconcepts.pp)
12. Global Measures of Spatial Autocorrelation: Moran’s I and Geary’s C (10SAglobal.ppt)
Week 5(Fall-2nd week)
13. Local measures of Spatial Association: Anselin’s LISA & others (11SALocal.ppt)
14. Using GeoDA for measuring spatial autocorrelation (12SADemo.pp)
15. Regression and Correlation: standard approach (13SpatReg.pp)
Week 6 (Fall-3rd week)
16. Spatial Regression
17. Using GeoDA for standard and spatial regression (15Regdemo.pp)
18. Analyzing surfaces: Concepts and Trend Surface Analysis (18Surfaces.pp)
Week 7 (Fall-4th week)
19. Analyzing surfaces: IDW and Kriging
18. Research at UT-Dallas: spatial statistics in practice. (16UTDRes.ppt)
19. A taste of other topics: cluster creation, network analysis, interaction data
26
Software(training will be included)
ArcGIS from ESRI, Inc. (licensed) GeoDA by Luc Anselin, Arizona State University (freely
available)
http://geodacenter.asu.edu/software/downloads Crime Stat III by Ned Levine, National Institute for
Criminal Justice (freely available)http://www.icpsr.umich.edu/CrimeStat/download.html
Copies also on my web site at:
www.utdallas.edu/~briggs
27
Research projects:what must be included.
28
US students seldom get this right.You must do better!!!
29
The seven parts of every Research Project Objective, explains the purpose of the research and why it is
important– Usually includes Hypotheses: possible explanations which you will test
Literature Review, identifies the key pieces of existing research relevant to the project and the hypotheses
Data Sources, identifies and explains the data used. Analysis and Methodology, explains the methodology applied to the
data. Results, describes the main research findings, whether or not the
hypotheses were upheld, and any potential problems or shortcomings Conclusion, discusses the implications of your results relative to your
initial project objective. References, provide standard format citations for all resources used
in the project. For more detail, go to:http://www.utdallas.edu/~briggs/poec6389/gisc6389_contents.doc
• These 7 parts are necessary for every research project
• Every research paper or report should include a section covering each part
30
Which component do you spend the most time on in your courses?
–Data and methods!! Which component is the most important for
successful research?
–The objective and the hypotheses
Objective: the foundation for any research project
31
Objective
Everything depends on the objective Hypotheses are based upon the objective data is selected which enables you to test the
hypotheses and meet your objective methods of analysis are applied to the data in order to
meet your objective Your results and conclusions are always made relative
to your objective
Everything links back to the objective!
32
Objective: one sentenceExample (project you can do): This research is in the area of environmental justice. Environmental justice is concerned with whether or not
some groups of people are more exposed to environmental hazards than other groups
“The objective of this research is to examine if low income or ethnic minorities in Dallas, Texas are more exposed to pollution than higher income groups or non-minorities”
33
Objective—your guide for everything!Your guide for every research decision What hypothesis will meet my objective? What data is needed to meet my objective? Which analytical method(s) is best to meet
my objective? How are my results relevant to my
objective?
34
If it doesn’t relate to the objective, don’t do it!
35
The Literature Review All research and analysis should build
upon the existing base of knowledge Must use the existing base of knowledge to:
– Identify appropriate objectives
– State logical hypotheses
– Select and use correct methodologies and appropriate data
All of this is accomplished by reviewing the existing literature
of scientific knowledge published in refereed journals not a Baidu or Google search
Base of Knowledge
Help with Literature ReviewsA standard Baidu or Google search is not sufficient! Instead, use Google Scholar • http://scholar.google.com/
The single most commonly used academic bibliographic database in US is probably “Web of Science” at
• http://isi10.isiknowledge.com/portal.cgi?DestApp=WOS&Func=Frame• Your library may provide access
A very useful bibliography fro GIS from ESRI is• http://training.esri.com/campus/library/index.cfm
Lists of GIS journals are available • On website of Dr Fang Qiu
– http://www.utdallas.edu/~ffqiu/• On my web site (with other information about GIS)
– http://www.utdallas.edu/~briggs/other_gis.html
36
How to Format Citations• It is very important to use the correct format when citing literature
– copying/pasting a URL is not sufficient!– It may be gone tomorrow
• The Chicago Manual of Style is the accepted norm.The Chicago Manual of Style. Chicago: University of Chicago Press, 15th
ed., 2003 – Or, copy the format used by any major journal in your field
• Good summaries of correct formatting are available at:– http://www.libs.uga.edu/ref/chicago.html – Or ,
http://library.osu.edu/help/research-strategies/cite-references/chicago-author-date/
• References are usually listed in a section at the end of the paper– For powerpoint presentations, it is good also to include the full citation
on the slides where it is discussed– A person reading a paper can "flip to the end" to check a reference, but
a listener cannot do that with a presentation!
37
Texts
O’Sullivan, David and David Unwin, 2010. Geographic Information Analysis. Hoboken, NJ: John Wiley, 2nd ed.
Other Useful Books: Mitchell, Andy 2005. ESRI Guide to GIS Analysis Volume 2: Spatial Measurement & Statistics.
Redlands, CA: ESRI Press.Allen, David W 2009. GIS Tutorial II: Spatial Analysis Workbook. Redlands, CA: ESRI Press.Wong, David W.S. and Jay Lee 2005. Statistical Analysis of Geographic Information. Hoboken,
NJ: John Wiley, 2nd ed.
38
Example ProjectYou do it!
Hypothesis Testing using ArcMap
39
www.utdallas.edu/~briggs
40
ArcMap Project
Zip file (proj1_data.zip)
Word document (proj1.doc)
41
ArcMap project
Purpose: Example of spatial analysis with objective,
hypotheses, etc. Experience with US data
42