Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

I am keenly interested in geo statistics (and many other words with 3 t's :).
I am not very programatically inclined but have been following several lists with this theme.
One being the R filter on stackexchange (I get occasional digest emails of questions there).
I find that there is an extremely intricate language associated with statistics in general.
I am insterested in learning from a layperson point of view what some of this language means and how to apply it to everyday GIS.

Please list any currrent sources for geo-statistics knowledge/understanding.

Here is an example of useful information...
travel for training
live online opportunity

Edit: Per request; I enjoy online training but get more from in classroom or step by step book (or pdf).

share|improve this question
    
Do you want to focus on Internet sources? Web pages? Blogs? Online courses? Academic conferences? Schools? Books? Case studies? etc., etc. –  whuber Jul 14 '11 at 17:40
    
In classroom is a bit too locational specific. What else can we say besides check out if a nearby university has applicable classes? Can you limit the domain of topical interests you have as well? –  Andy W Jul 14 '11 at 18:34
    
I think the domain is limited by geo-statistics. Is that what you mean? Also I have traveled 1000 miles for in classroom training in the past. –  Brad Nesom Jul 14 '11 at 19:51
    
This url is a re-broadcast of a geostat class given on FOSS not esri software. I recently noticed another live event upcoming. geostat-course.org/Landau_2011 –  Brad Nesom Oct 14 '11 at 19:32
    
ArcGIS would be prefereable, but essentially does not matter as the techniques cane be applied across various software. But in a perfect world I guess a lot of people are fimiliar with ArcGIS which makes the training easier, concentrate on the stats instead of learning which buttons to press in the software. Thanks for the link as that didn't come up on related posts. –  Rob Oct 14 '11 at 20:23
show 1 more comment

6 Answers

up vote 10 down vote accepted

This book, A Practical Guide to Geostatistical Mapping By Tomislav Hengl, is based on a 5 day course and uses open source applciations for all of the analysis, mostly R, SAGA, and Google Earth. You can download the book for free.

share|improve this answer
    
Thanks for providing the name of book. Would you mind to provide the active link where from this book can be downloaded. Thanks –  MYaseen208 Jun 19 '13 at 12:46
add comment

There is an excellent text for using R spatial statistic packages, including a chapter on Geostatistics.

Applied Spatial Data Analysis

share|improve this answer
    
@ dslamb - great link! –  Brad Nesom Jul 14 '11 at 18:13
    
In fact Roger Bivand is one of the instructors at the summer school referenced in my question. +1 –  Brad Nesom Jul 14 '11 at 20:24
add comment

For those with a similar lack of focus I would suggest perusing the listings of the GIS and Science blog. It is basically just listings of various research endeavors that have some relationship to geographical analysis, and so should qualify as "I am interested in learning from a layperson point of view what some of this language means and how to apply it to everyday GIS."


I most frequently encounter the adjective of geostatistics in conjunction with data analysis in the natural/environmental sciences. Examples of this are the texts Cressie (1993) or Isaaks and Srivastava (1989).

It is used far less frequently with statistical techniques more common in the social sciences. Examples of frequently cited texts focused on statistical analysis in the social sciences (but with an obvious focus on geography) are Anselin (1988), Waller and Gotway (2004), Lesage and Pace (2009), Ward and Gleditsch (2007). Books that might be considered a good bridge between the two fields might be Haining (2003) and Ripley (2004) (as well as the Bivand book cited by dslamb).

I list these because I don't necessarily endorse the distinction between the two fields (how can Moran's I not be considered a geostatistic?) But that being said, most people won't be particularly interested in all of those topical domains. Partly the reason that the distinction exists has to do with the type of data the statistical techniques are applied to, and hence if you are specifically interested in analyzing topical materials that are on one side the other may not be all that applicable. This is also the reason I suggested the GIS and Science blog, as they have listings falling under both of those categories. Although my interests largely remain in the social science realm, I still see articles more oriented towards the natural sciences that I find interesting (such as Visual Comparison of Moving Window Kriging models, now thats just cool!)

Now that I have inundated you with a plethora of expensive textbooks, are you still interested in all of geostatistics, or would your interests perhaps be slightly smaller in scope?

I frequently find that looking within software manuals are good places for definitions (and sometimes broader examples of applications). For example I came across the PASSaGE software when I was looking up a formula for local Geary's c. The GeoDa workbook is a wonderful introduction to spatial regression, and I've been told the manual/tutorials for the ClusterSeer software is a good introduction to cluster analysis (although unfortunately they do not have it available online it appears). For point pattern analysis CrimeStat is a very good reference.

Since I can imagine that learning the material in course format as opposed to a book is easier for some, I might suggest checking out if one of Pierre Goovaerts short courses on environmental geostatistics are coming nearby, and I see ICPSR has two courses related to spatial econometrics listed on their site (1,2, as a note these links will likely become outdated in the fairly near future). For entirely online material (and those of us who are more frugal), you can peruse the listings of MIT's open courses or for applied analysis using the R software you can work your way through the spatstat tutorial.

Also since travelling 1000 miles for a course is rarely feasible, if you find a course that looks interesting asking the professor for a copy of a syllabus is a good way to identify pertinent reading material. There was recently a post on the stats site asking for software recommendations to estimate variograms, and I would think it is likely there are some more useful sources of learning material listed on that thread.

Just to continue on rambling with resources I have collated, besides the Hengl (2009) book that was already listed in your question, below are other websites with various resources;

  • CATMOG (A note, these are a good place to start for an introduction to the specific topical material that is covered)
  • Geospatial Analysis - A comprehensive guide (de Smith, Longley, and Goodchild, 2006) which I am sure has been cited here multiple times.
  • The Center for Spatially Integrated Social Science has a host of resources.
  • For resources related to visualization I have found GeoVista, and the Spatial Data Mining and Visual Analytics lab to have some pretty cool stuff.
  • The resources at the Geoda center are worth mentioning a second time (although they could perhaps use some better organization!) @Laurent mentions the tutorials page, which has some software tutorials for spatial regression, point pattern analysis, and variography in different software packages. I was recently forwarded a page of e-presentations from them as well. It is probably the most wide variety of spatial analysis presentations I have ever seen, spanning the divide between natural and social science techniques I discussed earlier in the post. I haven't gone through the slides, but I highly suspect they are a good introduction to any of the topics they cover (and likely less intimidating an introduction than from some of the text-books I listed prior). I find new things on that site every time I peruse it, it is worth mucking around to see if I missed anything.
share|improve this answer
    
I like it; but any suggestions for finding what I need on that site? archives, categories not sure how to search. How do you use it? Just watching as blogs roll by? –  Brad Nesom Jul 14 '11 at 18:48
1  
Practically all of the research listings fall generically under the category of "geostatistics" on that site. I personally just signed up for the RSS feed and then if an article I'm interested in is posted I read it (I'm not quite sure how else you are supposed to use it). I don't see a way to search either (besides the drop-down categories). –  Andy W Jul 14 '11 at 18:57
    
thanks that's good. –  Brad Nesom Jul 14 '11 at 19:06
2  
As a note to anyone reading this, I am well aware that this is entirely gratuitous and no one should take this daunting list as a must read to become saavy within the domain of geo-statistics. I did it just as much for myself to organize all of the bookmarks I have collected in an easily readable format! –  Andy W Jul 15 '11 at 13:24
    
Excellent @Andy W - good edit! –  Brad Nesom Jul 15 '11 at 14:18
add comment

I second the mention of the GeoDa workbook by Andy W., in fact there are other interesting learning ressources on the "Documentation" page of the Open GeoDa project.

share|improve this answer
add comment

Have you been through the esri geostatistical analyst resources?
Esri Web courses
Esri Instructor Led Training
Esri Geostatistical Analyst Help
Esri Geostatistical Analyst Tutorial

share|improve this answer
1  
Short of giving an answer with the correct format and software... There are great solutions and resources on this list. stat.ethz.ch/mailman/listinfo/r-sig-geo –  Brad Nesom Oct 14 '11 at 21:41
add comment

This is a very good community of people interested in Geospatial data and statistics with a focus on Geostatistics.

You will find course materials, wiki pages, important links related to spatial statistics and geostatistics etc.

Spatial analyst net

You should check the following R packages

  1. gstat
  2. automap

Gstat is a comprehensive Geostatistical library and automap library helps you run kriging method relatively easily

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.