UCGIS Virtual Seminar - Fall 1998 [Back][Refresh][Options][Search] Spatial Analysis in a GIS Environment [Edit*][Delete*] [Image] Spatial Analysis in a GIS Environment art getis 09/14/98 [Image] [Image] One way to find this white paper Dawn Wright 09/14/98 is to go the UCGIS Virtual ... [Image] This is a test.... Guangxiang 09/14/98 Cheng [Image] Spatial Analysis Article Erik Shepard 09/14/98 [Image] [Image] User oriented art getis 09/15/98 [Image] Spatial Analysis and GIS Byong-Woon Jun 09/14/98 [Image] Visualization is key Wilmot Greene 09/15/98 [Image] [Image] re:visualization and sound Ronald William 09/18/98 Ward [Image] Data Availability, Analytical Capacity, Bill Moseley 09/16/98 and Decision Making [Image] [Image] GIS and spatial analysis Erik Shepard 09/16/98 curriculum [Image] [Image] Erik, Our complaint all along art getis 09/18/98 has been that GIS technolog... [Image] [Image] Bill makes good points. As has art getis 09/18/98 always been the case, there ... [Image] [Image] Towards User-friendly Spatial Byong-Woon Jun 09/18/98 Analysis in GIS [Image] [Image] Problems with user friendly Erik Shepard 09/18/98 systems [Image] [Image] Re:Problems with user Byong-Woon Jun 09/18/98 friendly systems [Image] [Image] Integrating GIS and Erik Shepard 09/19/98 analysis curriculum [Image] [Image] Comments:Integrating GIS Byong-Woon Jun 09/19/98 and Spatial Analysis Curriculum [Image] [Image] I agree that we will Erik Shepard 09/21/98 not quickly move towards this kind of r... [Image] [Image] assumptions about Ronald William 09/22/98 policy maker's statistical knowledge Ward [Image] [Image] The seductive allure of GIS Bill Moseley 09/22/98 [Image] [Image] GIS democratization & chris watson 09/23/98 misinformation [Image] [Image] Chris makes good points Bill Moseley 09/29/98 [Image] General comments Ronald William 09/16/98 Ward [Image] [Image] Background Caiming Shen 09/21/98 [Image] Philosophy behind analysis Ronald William 09/16/98 Ward [Image] [Image] General Comments Erik Shepard 09/16/98 [Image] [Image] point well taken Ronald William 09/16/98 Ward [Image] [Image] re: philosophy Ronald William 09/17/98 Ward [Image] [Image] Exploratory and Confirmatory Byong-Woon Jun 09/17/98 Spatial Data Analysis in GISs [Image] [Image] Agreed Wilmot Greene 09/18/98 [Image] Introduction Jimmy Knudsen 09/16/98 [Image] [Image] Welcome Jimmy!! Dawn (Oregon Dawn Wright 09/16/98 State)... [Image] Ambiguous Term Definition in the White Byong-Woon Jun 09/17/98 Paper [Image] [Image] spatail analysis vs spatial data Ronald William 09/18/98 analysis Ward [Image] [Image] Hellow from a "veteran" of the xiaojun yang 09/18/98 virtural seminar! [Image] [Image] Hey Ron, Good question! "Should Wilmot Greene 09/21/98 I be thinking of spat... [Image] [Image] Inductive vs Dedutive Approach Byong-Woon Jun 09/22/98 in a GIS Environment [Image] People and Pixels: Oil and Water? Molly Brown 09/19/98 Cisse [Image] [Image] Paper by Parsons Wilmot Greene 09/20/98 [Image] [Image] that reference... Molly Brown 09/21/98 Cisse [Image] [Image] Analysis of human-environment Bill Moseley 09/22/98 relationships [Image] [Image] Bill, There are several art getis 09/22/98 other white papers in the UCGIS s... [Image] [Image] Human environmental Ronald William 09/23/98 relationships and profit Ward [Image] [Image] Spatial data acquisition and Byong-Woon Jun 09/22/98 integration , and Scale [Image] [Image] Thanks! Molly Brown 09/23/98 Cisse [Image] [Image] PCV? Ronald William 09/23/98 Ward [Image] [Image] Yes, of course... Molly Brown 09/30/98 Cisse [Image] Greetings Jay Raiford 09/20/98 [Image] [Image] Re:Greetings Byong-Woon Jun 09/20/98 [Image] [Image] Re: Greetings Jay Raiford 09/21/98 [Image] Spatial Analysis in a GIS Environment Kurt L. Johnson 09/20/98 [Image] More research building towards a better Esra Ozdenerol 09/21/98 established framework [Image] Summary of Digital Earth and 5 research Esra Ozdenerol 09/21/98 priorities [Image] Discuss of digital earth and other five Guangxiang 09/21/98 research priorities Cheng [Image] Digital Earth art getis 09/21/98 [Image] [Image] comments on the proposed 1999 Ronald William 09/22/98 white paper on 'digital earth.' Ward [Image] [Image] Comments on the Proposed 1999 Byong-Woon Jun 09/22/98 White Paper on The Digital Earth [Image] [Image] social theory and GIS Ronald William 09/25/98 Ward [Image] [Image] Re:Social Theory and GIS Byong-Woon Jun 09/25/98 [Image] Background Caiming Shen 09/21/98 [Image] comments on spatial analysis Guangxiang 09/21/98 Cheng [Image] metadata, spatial data analysis, and Nina Lam 09/21/98 others [Image] [Image] clarification request Ronald William 09/22/98 Ward [Image] [Image] posting writeups Dawn Wright 09/23/98 [Image] FYI: Al Gore's Speech on the Digital Byong-Woon Jun 09/21/98 Earth [Image] Digital Earth plus 5 topics not covered Jay Raiford 09/21/98 [Image] SOFTWARE, the Key of Spatial Analysis? xiaojun yang 09/22/98 [Image] Opinions and thoughts Erik Shepard 09/22/98 [Image] Question regarding distinct treatment of Bill Moseley 09/22/98 spatial data [Image] [Image] Re:Bill's Question Byong-Woon Jun 09/24/98 [Image] Progress in Spatial Analysis and Modeling Byong-Woon Jun 09/22/98 in a GIS Environment [Image] Request Jay Raiford 09/23/98 [Image] [Image] agreed - THIS IS IMPORTANT!!! Ronald William 09/23/98 Ward [Image] Looking Toward the Future art getis 09/23/98 [Image] [Image] To me spatial analysis is GIS. We Jay Raiford 09/23/98 need, as students, to lea... [Image] [Image] Re: Looking toward the future Erik Shepard 09/24/98 [Image] [Image] Re: Looking Toward the Future Byong-Woon Jun 09/24/98 [Image] [Image] re: looking toward the future Ronald William 09/25/98 Ward [Image] [Image] General Opinion:Looking Toward Mihye Bark 10/05/98 the Future [Image] [Image] U.S. at the cutting edge? Doug Albert 10/05/98 [Image] [Image] Answer four questions Guangxiang 10/07/98 Cheng [Image] Report for Spatial Analysis in a GIS Mihye Bark 10/04/98 Environment [Image] Spatial Analysis Report - LSU Jay Raiford 10/05/98 [Image] Spatial Analysis in GIS 2nd try... Kurt L. Johnson 10/05/98 [Image] Summary of digital earth and five Doug Albert 10/05/98 research priorities [Image] Spatial Analysis in a GIS Environment Doug Albert 10/05/98 [Image] Spatial Analysis Report Guangxiang 10/05/98 Cheng [Image] ... Guangxiang 10/05/98 Cheng [Image] Post new message in this thread ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 14, 1998 09:43 AM Author: art getis (arthur.getis@sdsu.edu) Subject: Spatial Analysis in a GIS Environment Hello. Is this message getting through to you? Please begin the seminar by reading the WHITE PAPER (produced in 1996) that can be found in Research/White Papers/Spatial Analysis in a GIS Environment. If any questions come up while you are reading it, please post them. I encourage seminar members to attempt to answer questions as well as raise them. Currently, the white paper is being up-dated; so keep in mind that it represents thinking in the historical past. Welcome aboard. Art Getis (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1151) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 14, 1998 02:01 PM Author: Dawn Wright (dawn@dusk.geo.orst.edu) One way to find this white paper is to go the UCGIS Virtual Seminar course web site (http://dusk.geo.orst.edu/virtual), and click on the schedule. There is a link under "Spatial Analysis." I believe that this is the 1996 paper that Art is referring to. Dawn (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1159) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 14, 1998 12:06 PM Author: Guangxiang Cheng (gcheng1@tiger.lsu.edu) This is a test. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1155) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 14, 1998 07:49 PM Author: Erik Shepard (shepard@uga.edu) Subject: Spatial Analysis Article My first thoughts on reading this article is that the basic gist is really not new (Dr. Getis did mention that the article represented the historical past). I remember back when I took my first GIS class hearing about the need to integrate spatial analysis and GIS. I do think, though, that the mention at that time was more of a desire rather than a direction; the hardware available at that time simply did not support the numerically intensive operations required to use analytical techniques. These days, with the proliferation of 300+ megahertz machines this is not quite the same obstacle (although with the advent of finer and finer resolutions of data we continue to "push the envelope" in terms of machine capabilities). I do think that the article raises some particularly valid points about incorporating time series data as well as very large databases. Again, with time series data, there was always the desire but we need now to come up with models which will allow us to track this sort of information (4+ dimensional databases?). With the recent explosive growth of spatial data aquistion, large databases have also become a serious issue. Regarding the formalization and enhancement of statistical techniques, this is not particularly new either, but is something that again we can finally begin to work on with current hardware support. Overall, my impression was that while the information contained in the article was not shockingly new, it was nice to finally see a formalized direction. Now that we have the hardware, we really should start putting it through its paces to really use GIS to solve some problems. That's what GIS is good for, and I know that's one of the things that originally attracted me to it. Opinions and comments, anyone? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1175) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 15, 1998 11:05 AM Author: art getis (arthur.getis@sdsu.edu) Subject: User oriented It almost feels as though Erik atttended the UCGIS meeting of the Spatial Analysis group in Park City last June. At that time we identified those areas that seemed to us to be the next steps in spatial analysis in a GIS environment. Primarily, we focused on the academic user community. It is time, we said, that our analytical devices, in a user friendly format, should be made available to those practitioners who are trying to solve real world problems. In fact, we changed the name of the focus to Spatial Analysis AND MODELING in a GIS Environment. Of course, we were conscious of new developments in geo-computation and visualization and how these can help the analyst. So the emphasis, in general, was toward measurement, user useful software, large data sets, and new technology. Thanks, Erik, for describing it so well. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1182) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 14, 1998 09:17 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Spatial Analysis and GIS According to the white paper, I think it shoud be clear that, influenced by GIS, RS, the availability of very large, high spatial resolution data, and access to extremely computer power, spatial (data) analysis has already changed greatly and will continue to change as methods which realize the existence of today's data and computer-rich environment are developed. On the other way, current GISs suffer from a variety of problems such as poor performance for many analytical operators, poor ability to handle dynamic spatial models, and poor handling of the temporal dimension. To solve these problems, a need has existed to integrate spatial analysis and dyanmic modeling with GISs. Actually, the mutual benefits(for examples, visualization of data by means of GIS and advanced analytical methods from statistics) of closer links between GISs and spatial (data) analysis have offered such an integration. I think the emphasis of the paper is more upon developing new conceptual framework for spatial analysis in new computing environment. Probably, the next step may be application-oriented(?). (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1177) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 15, 1998 05:59 PM Author: Wilmot Greene (mot@uga.edu) Subject: Visualization is key The aspect of analysis that interests me the most is visualization. I am currently researching how the use of sound can be incorporated in aiding cartographic visualization. There is little or no use of sound (as far as I know) in GIS software. Analysis of data is limited by the analyst. Therefore, we should focus on how to increase visualization. By incorporating sound in a GIS environment it would be possible to simultaneously analyze many data for any area, which would increase the analyst's understanding of the relationships between those data. Some research that I have come across in my literature review suggests that sound is particularly useful with time series and multi dimensional data. Let me HEAR your opinion. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1185) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 10:42 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: re:visualization and sound I think the discussion group should be aware of some of the applications for sound Wilmot Greene has mentioned to our group here at UGA. One such application Wilmot mentioned is the use of sound for searching large data bases. Visually speaking, reading through tables of information or looking through large numbers of time series maps may be impractical. However, if one could combine a rapid data search engine with a sound identification capability, to ID statistically significant overlays with a tone (when there might be thousands of such overlays to construct - most of which might not mean anything), then huge image data sets could be analyzed much more efficiently than they are now. This is exciting. Ron Ward (thanks to Mot Greene) (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1225) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 04:27 AM Author: Bill Moseley (wmoseley@uga.edu) Subject: Data Availability, Analytical Capacity, and Decision Making A major theme of the white paper seems to be that GIS analytical capacity has not kept pace with the volume of data available. Furthermore, as Dr. Getis mentioned in his last message, we are trying to make this improved analytical capacity available to practitioners and decison makers. As analytical processes improve and GIS continues to democratize, is there a risk that GIS neophytes (like myself) will have the power to draw conclusions without fully understanding the assumptions of the analytical model employed? While I think increasing access to increasingly powerful models and larger datasets is a good thing, academics (who presumably understand the assumptions that have been made) cannot simply "make available" the tools but, perhaps, have a responsibility to stay involved in the political, decision-making process. What do they say, "with power comes responsibility." I look forward to reading any reactions. Cheers! (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1188) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 07:05 AM Author: Erik Shepard (shepard@uga.edu) Subject: GIS and spatial analysis curriculum I think that Bill makes a good point about not only being able to do spatial analysis with GIS but also understanding the implications of what you have done as well. It brings me to something that I have been thinking about, which is that the paper indicates research directions of tying in spatial analysis to GIS to be able to use GIS to perform these sorts of things. But shouldn't we also consider the flip side, which is trying to integrate GIS into current geostatistics curriculum? Professors of spatial analysis today (at least in my experience) have no qualms about using packages like SAS and SPSS, but perhaps we need to start looking at integrating the use of GIS into these courses as well (emphasizing not only the applications but also the theory). It seems like this would help to alleviate some of Bill's concerns about misuse of GIS for decision making. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1189) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 10:51 AM Author: art getis (arthur.getis@sdsu.edu) Erik, Our complaint all along has been that GIS technology has not gone strongly in the direction of data analysis. In the last couple of years, however, we are beginning to see some movement in this direction. For example, at one time one couldn't do geostatistical operations in a GIS, but now some items like semivariance creation and kriging are included in some GIS packages. I find, however, that these routines are inferior to some stat packages that I have come across. I hope that the trend toward making GIS more sophisticated analytically would continue, and speed up. Art (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1227) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 10:44 AM Author: art getis (arthur.getis@sdsu.edu) Bill makes good points. As has always been the case, there are several control points along the way. These include instructors, editors, and referees. If these people are aware of the assumptions inherent in the data analysis routines, and rigorously defend the principles of proper specification, proper estimation, clarity, data availability (for checking), and proper answers to questions about procedures, then there is little need to worry. Unfortunately, however, not all of these people are well schooled in the nature of the proper use of the methods employed. At the very least, as students and researchers, we should constantly ask ourselves if we know what we're doing every time we 'point and click.' (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1226) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 01:57 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Towards User-friendly Spatial Analysis in GIS There are several user inferface methods available. I think one of them is to utilize online help menu in a GIS in the short term. For example, if you carefully take a look at IDRISI, it provides user-friendly online help menu for us which explains command usages and their analytical operation procedure. We can easily use spatial autocorrelation function in IDRISI even though we assume we are GIS neophytes and have no idea of spatial autocorrelation. I think most of GIS software provides such a user friendly online help menu in these days. Don't you think that the user-friendly help menu in a GIS program which explains command usages and their analytical operation procedures is a way to alleviate misuse of GIS for decision making in the short term? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1230) ------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 02:03 PM Author: Erik Shepard (shepard@uga.edu) Subject: Problems with user friendly systems The problem that I see with this is that although there may be online help which may even explain and detail some of the assumptions of a statistical technique, it does not mean that someone reading it will understand and properly apply these assumptions. In some ways, I think, user-friendly analysis systems worsen the problem of "misuse" because anyone can fill in a couple of fields with some numbers and submit the problem without ever really knowing what they are doing. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1231) -------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*] [Delete*] Date: September 18, 1998 02:42 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Re:Problems with user friendly systems Maybe you are right. If so, what's the best solution to this problem? Education or training people invloved with decision making. As Dr. Getis indicated, things haven't come up with our desire. What if someone related with decision making has no idea of computer and GIS, but he/she knows much of statistical techniques? What will happen? How can we make user-oriented computer or GIS? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1233) --------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*] [Delete*] Date: September 19, 1998 01:38 PM Author: Erik Shepard (shepard@uga.edu) Subject: Integrating GIS and analysis curriculum I think that part of the solution lies in not only offering some training in analysis in GIS courses, but also in offering some training in GIS in analysis courses. There needs, I believe, to be some reciprocity in order for this integration to truly take place. As I said before, we regularly are taught or teach software such as SAS and SPSS in our spatial analysis classes. Why not teach Arcview or Idrisi as well to illustrate some of the spatial points such as kriging or spatial autocorrelation? I don't think that we can guarantee that policy makers are versed in spatial analysis techniques any more than we can stop anyone with Microsoft Excel and some data from running regression analysis without really knowing what it means. I do think that by bridging the gap here and offering an integration not only from the GIS side but also from the courses in analysis, we begin to deal with the problem. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1268) ---------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*] [Move*][Delete*] Date: September 19, 1998 07:46 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Comments:Integrating GIS and Spatial Analysis Curriculum As Dr. Getis indicated above, "GIS community has not gone strongly in direction of spatial analysis. However, we are beginning to see some movement in this trend". Again, things have not always caught up with our desire. Probably in the not near future, we might do what you insisted. However, I do think that a variety of user-friendly interfaces would help bridge the gaps between spatial analysis and GIS by now. And we firstly need to develop more sophisticated analytical modules in GISs before developing such user interfaces. As of now, there are small number of statistical analytical functionalities included in some GIS packakges as Dr. Getis mentioned. I'm wondering to what degree this approach can be implemented in current commercial GISs which are thoroughly market-oreinted. Do you think that teaching GIS software (IDRISI and ArcView) in an integrated statistical class will help us know how map algebra (a kind of analytical operator) really works inside GISs? I don't think so. Finally, please don't underestimate GIS users' or policy makers' analytical abilities. How can you easily estimate other's analytical intelligences based on such an stereotype about decision makers? Who knows their abilities? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1274) ----------------------------------------- [Top][Previous][Next][Print][Reply] [Edit*][Move*][Delete*] Date: September 21, 1998 05:55 AM Author: Erik Shepard (shepard@uga.edu) I agree that we will not quickly move towards this kind of reciprocity. However, I don't think that means we shouldn't even try. I also agree that developing more tools in the software will facilitate this kind of cooperation, but I don't think that that is the only solution either. In regards to your question about GIS software in an integrated statistical class, no absolutely not. Teaching GIS will not teach spatial analysis theory any more than teaching a student how to use SAS (which we do today) will teach them to understand regression. What is does do is present them with a theoretical overview a la classwork and then present them with tools in the software (e.g. the GIS when we develop these tools) to use these theories. And finally, please understand that I do not assume that policy makers don't know about statistical techniques. I never said that they didn't. What I did say was that I was concerned that someone without an understanding of the techniques (and this could just as easily apply to myself or to you as to policy makers) would be able to easily misuse them given the kind of user interface improvement that we are talking about here. I simply endorse not only integration of tools into the software but also the educational background to properly use them. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1298) ----------------------------------------- [Top][Previous][Next][Print][Reply] [Edit*][Move*][Delete*] Date: September 22, 1998 07:01 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: assumptions about policy maker's statistical knowledge In reply to Jun's observation that it is difficult to assess the statistical abilities of policy makers and GIS users: What is the least common denominator here? I would rather assume that many GIS users do not have a working knowledge of statistical assumption, rather than assuming that they do. To not do so is dangerous. For example, in the review process aothors go through before having a piece of work go to publishing we do not assume that everything written is sound reasoning, but instead, we leave it to reviewers to assess if the research is sound. Being sceptical about the statistical abilities of GIS users is much the same. If we assume that all reasoning involved in any spatial analysis is sound, we run the risk of allowing misinformation to become policy, even though there might be serious problems with the methodologies therein. I'm on Erik's side here...putting too much faith in statistics, even when all the assumptions are met, is dangerous. Keep thinking critically! R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1347) -------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*] [Delete*] Date: September 22, 1998 08:41 AM Author: Bill Moseley (wmoseley@uga.edu) Subject: The seductive allure of GIS Eric and Byong-Woon: I think making sophisticated analysis capacity commercially available is a good thing. The assumptions involved in statistical analysis will always be inaccessible to the majority of the American public. The problem is that while statistical results are drab and dreary, GIS output can be pretty damned sexy. That's right, policy makers (Al Gore?) and pseudo GIS wonks like myself have been seduced by the allure of GIS because maps are a great mode of communication. The challenge is not only to make analysis more accessible, but to be as up front and user freindly with the assumptions as possible (and that means doing more than burying the assumptions in a help menu). Cheers! (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1352) --------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*] [Delete*] Date: September 23, 1998 07:26 AM Author: chris watson (watsonce@arches.uga.edu) Subject: GIS democratization & misinformation Just a couple of comments regarding the previous statements by Bill and Ron. The two major concerns being addressed seem to be 1) making GIS user friendly and available to the masses, while 2) addressing concerns about misuse by policy makers (and others) based on misunderstanding of analytical/statistical procedures and assumptions. My gut feeling is that by realizing 1), then 2) will be taken care of, although not immediately. This can be seen in the increasing use by environmental groups (especially small , locally based ones) of GIS to not only examine and challenge policy and management decisions by state and federal agencies, but to also generate alternative management plans (the acceptance of which is a whole different battle). My point is that once the little people get their hands on the technology, they will not hesitate to challenge the policy makers. At the beginning stages of this process the understanding of analytical procedures may be limited (on all sides), however over time sophistication is bound to increase simply due to the dialectical nature, not to mention the contentiousness, of the political process. Based on this line of thinking, I feel that making the technology accessible, should override concern about misuse. Then again, I could be wrong. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1380) ---------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*] [Move*][Delete*] Date: September 29, 1998 11:22 AM Author: Bill Moseley (wmoseley@uga.edu) Subject: Chris makes good points Chris: I think you make some very good points. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1580) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 07:18 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: General comments Hey folks...good to see we finall have the ball rolling. A couple of general comments here: E. Sheppard here at UGA makes the observation that handling data has become easier because of the proliferation of more powerful PCs...but isn't part of the point of this White Paper just the opposite - we don't really have the capability to carry out analyses on large time-space data sets. Another important point of the article is that we need to develop new statistical techniques to analyze time-space data. Can we assume that statistical methods of analysis currently in use are appropriate for analyzing large time-space data sets? I'm confused on this point. Bill Mosely brings up another important point: many of the statistical analysis techniques included in GIS software require that the user can manipulate the software, but do not necessarily require that the user understands the assumptions inherent in statistical methodology. I guess the same is true of statistical analysis software packages like SAS and the like, but to read output tables and graphs produced by these statistical software packages one must have a good feel for what is going on statistically, or else the user won't know what type of analyses to use on any particular data set in the first place. GIS spatial analyses are different for one important reason: map representations of spatial analyses do not usually include insets of raw data or univariate tests for normality and other such information. So how do we know there are sound statisticqal principles behind every IDRISI or ARC-INFO map we see? A good example of this problem can be seen in computer cartography. Years ago, before computer mapping software, if you wanted to learn cartography you had to learn good cartographic design and objective representation principles at the same time. Now, persons ignorant of cartographic principles can learn Atlas GIS or Maptitude without necessarily having learned any of the good science behind good, objective cartography. Witness the slew of poorly designed maps we are seeing in newspapers and magazines across the country (for those of you here at UGA the maps in the Red and Black are evidence enough). Design, however, is an aesthetic state of affairs which may be neither here nor there. The real problem is that many computer generated maps may contain subjective information. For example, if somebody wants to generate a choropleth map of the U.S. population by county, there are any number of ways to sujectively set data classes so as to show areas of dense or sparse population where you want them to be (say for the case of congressional redistricting). John Q. Public at large has little understanding of how cartographic representations can be used for propoganda, and will read the map and make political decissions based on what he sees. So, the responsibility for truthful representation is with the cartographer...yet cartographic software is user-friendly to persons other than objective scientists. Data can be manipulated within the GIS environment, and like any statistic, spatial data representations are only as truthful as the producer. I guess objective statistical analysis (or lack thereof) is not an issue unique to GIS, but I'm not reading any coverage (excuse the pun) of the issue in this first White Paper. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1190) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 10:35 AM Author: Caiming Shen (cmshen@hotmail.com) Subject: Background I agree withI agree with ward. The white paper emphasizes the capability to handle large spatial and spatial-temporal data sets. However, I wonder whether this question is related to hardware or methods of spatial analysis? After I read the white, I did not found that what the spatial analysis can do at present. My meaning is that we need a background which tell us what kinds of statistical methods are being used in current GIS software, what kinds of problems those methods can be used to solve, how great data sets they can handle, whether those methods are appropriate for handle large spatial and spatial-temporal data sets, how far the spatial analysis in GIS has gone, and so on. I think it is important, especially to newcomers like me. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1313) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 07:35 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: Philosophy behind analysis From the first White Paper... "enormous quantities of data are available to help solve local and regional problems. We must devote energy to exploit this availability." My understanding of good science is that first we formulate questions about a particulat phenomenon, then we develop hypotheses pertaining to these questions, then we develop sampling and analytical methodologies to test these hypotheses - and finally we collect data. In other words, our questions drive data collection. In this case we have data and we are trying to think of ways to analyze it for increased profitability (I'm uncomfortable at the thought of this) and or social benefit. Much of the critisism I read and hear about the use of GIS is that much of what goes on is 'data driven,' and that this leads to the observation that GIS is a solution looking for a problem. It is acceptable to take the inductive approach and thus allow the answers to emerge from the data analysis(as a biogeographer my understanding is that this is usually what we do)...but should we have questions in mind (not necessarily suspecting certain answers at the same time, which is the deductive approach, and still should be driven by questions) before we collect the data? Should we at least have questions in mind before seeking out a secondary data source (if utilization of exsisting data, being cost effective, is the issue)? R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1192) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 08:11 AM Author: Erik Shepard (shepard@uga.edu) Subject: General Comments To clarify my point a little bit; I do think that Ron is correct that while we may have the processing power we don't yet have the techniques to harness that power. What I was saying though, was simply that the idea that "we should develop some techniques to use GIS and spatial analysis" is not new, but perhaps the capability to finally achieve that goal is now attainable. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1194) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 08:15 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: point well taken Sorry for any misrepresentation, Erik. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1195) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 17, 1998 07:50 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: re: philosophy Yesterday in our one hour session, Lynne Usery was good enough to explaion the difference between exploratory (after Stan Openshaw) and confirmatory investigative methods. Confirmatory methods are those involving developing a hypothesis (questions) and methodology, then going on to collect data. Exploratory investigation, in principle, is searching through data in an attempt to see what can be gleaned from it, then questions arise and the usual confirmatory investigation can begin from that point. I used the analogy of Darwin, who explored for a period, formulated questions afterward, then went on with confirmatory methodology to arrive at what are now well known conclusions. With regard to GIS, these exploratory methods are virtual in that the exploration is carried out on data (as opposed to the physical exploration carried out by Darwin). In principle then, the exploratory path in GIS can lead to good science. Of course, there have been instances where this is not the case, which. I think, brings us back to Bill's original point: user-friendly GISs might generate poor science. Still, I accept the validity of the exploratory approach...providing it's based in sound scientific principle. Thanks for the clarification, Lynne! R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1207) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 17, 1998 12:17 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Exploratory and Confirmatory Spatial Data Analysis in GISs As we dicussed yesterday, spatial data analysis is divided into two major categories: exploratory and confirmatory spatial data analysis. First, "exploratory spatial data anslysis invloves seeking good descriptiions of data, thus helping the analyst to develop hypothesis about such data." Such techniques are charaterized by making few a priori assumptions about the data and many are designed specifically to be as resistant as possible to the effect of extreme data values (called outliers). It is necessary that many such methods emphasize graphical views of the data which are designed to highlight particular features and allow the analyst to detect pattern, relationships, usually values, and so on. The views that result from exploratory methods with GIS may be in the form of maps and others may include more conventional plots. In this regard, visualization techniques in GIS contribute to spatial analysis or spatial data analysis. Second, confirmatory spatial data analysis is to test and confirm the hypotheses which are formulated prior to any knowledge of the data to be tested. So far, establishing such a priori hypotheses clearly limits its usefulness in a GIS context even if the technical problems of doing it can be overcome. We need to develop the confirmatory (significance) procedures in GIS. However, several studies show that current GISs are more relevant to the exploratory spatial data analysis (ESDA). Any comment? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1208) ------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 09:30 AM Author: Wilmot Greene (mot@uga.edu) Subject: Agreed I agree with Jun. We have established the fact that data exist in almost unfathomable quantities. This amount of information provides reason enough to do exploratory GIS. And visualization techniques are a sound way to carry out this exploration. Visualization (shoes) >> Spatial data analysis (ability to run) >> Spatial analysis (winning a race) Mot (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1221) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 05:53 PM Author: Jimmy Knudsen (jknudsen@ycp.edu) Subject: Introduction Allow me to introduce myself. Hello, I am Jimmy. I accidently found this virtual forum while looking on the internet to learn about GIS. I am now taking a course in GIS at York College of Pa. It is called Maptitude. I have graduated from YCP with a degree in International Relations. My minor was and is geography and I wish to integrate the two and work in the Middle East. Allow me to just sit and read and learn from you and if I might ask some questions every now and then. Nice meeting everyone. See Ya! (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1201) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 16, 1998 06:09 PM Author: Dawn Wright (dawn@dusk.geo.orst.edu) Welcome Jimmy!! Dawn (Oregon State) (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1202) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 17, 1998 12:57 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Ambiguous Term Definition in the White Paper From the first page in the white paper: "The term "spatial analysis" encompasses a wide range of techniques for analyzing, computing, visualizing, simplifying, and theorizing about geographic data." I think the term distinction between spatial analysis and spatial data analysis in the white paper is unclear. My feeling is that the white paper generally used the term spatial analysis throughout pages in the light of spatial data analysis. We need to draw the particular distinction between spatial analysis and spatial data analysis. Spatial analysis is defined as the the quantitative study of phenomena that are located in geographic space to describe their pattern, to explain their process, and to predict their relationship. "Spatial data analysis involves the accurate description of data relating to a process operating in space, the exploration of patterns and relationships in such data, and the search for explanations of such patterns and relationship." Spatial analysis is broader concept than spatial data analysis. A good spatial data analysis helps facilitate a good spatial analysis. It is also important to distinguish one from another when we try to integrate spatial analysis or spatial data analysis with GISs. Again, I think the emphasis of the white paper is more towards spatial data analysis. Comments or opinions? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1209) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 10:31 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: spatail analysis vs spatial data analysis I agree with Jun: this first White Paper refers almost entirely to spatial analysis and makes little reference to spatial data analysis. Maybe those of you working within this thread (Jun and Mot) can clear up some of my confusion. Should I be thinking of spatial analysis in terms of being inherent to confirmatory research, and spatial data analysis, then, in terms of exploratory research? Am I trying to compartmentalize the four together into categories into which they do not necessarily belong? Also, can somebody give me a practical example of how metadata fit into this thread of discussion? My limited understanding is that metadata is data about data (the acquisition and integration paper outlined some standards that should be applied to metadata structure) - but what are the practical/applied uses of metadata? Is metadata an index necessary for exploratory research through large data bases? Can somebody give me a working example? Ron Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1224) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 18, 1998 12:57 PM Author: xiaojun yang (yang@uga.edu) Subject: Hellow from a "veteran" of the virtural seminar! Greetings! As a participant of last year's seminar, I am pleased to be able to access this site again. I notice this year's seminar is experiencing many changes! Ron and Jun are so active right now! Well done! Regards, Xiaojun Yang University of Georgia (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1229) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 09:21 AM Author: Wilmot Greene (mot@uga.edu) Hey Ron, Good question! "Should I be thinking of spatial analysis in terms of being inherent to confirmatory research, and spatial data analysis, then, in terms of exploratory research?" I think not. To me ( I certainly could be wrong ) spatial data analysis is being done when the analyst is looking for patterns and/or clues in a data set that disclose themselves via any method. While true spatial analysis requires a more comprehensive look at the landscape. This requires applying those patterns and/or clues with other patters (layers). It's almost like if you are analysing one layer you are doing spatial data analysis. But, if you are dealing with more layers you are doing spatial analysis. So it doesn't matter if you are looking for a problem (exploratory) or looking for a solution (explanatory). I have no literature to support these thoughts, that is just the way I understand it. As far as the Metadata question,,, I don't get it either. MOT (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1302) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 06:35 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Inductive vs Dedutive Approach in a GIS Environment 1. Comments on Ron's Question: First of all, I'd like to resummarize categories of spatial (data) analysis. Spatial (data) analysis is divided into two major categories: Confirmatory and Exploratory spatial (data) analysis. This means we can take spatial analysis or spatial data analysis in terms of either confirmatory or exploratory approach. In other words, there is no assumption that we should try to consider spatial analysis in terms of confirmatory approach and spatial data analysis in terms of exploratory approach. I hope it helps Ron better understand what I was talking about. 2. Inductive vs Deductive Approach in GISs: According to my literature review, we can take inductive or deductive approach in GISs. From the inductive approach, GIS is a data-rich environment and thus should be equipped with more robust geographical exploratory techniques. In this category, the researchers obviously favor more inductive analytical capabilities in GIS with incorporation of spatial statistical procdures. Specific techniques that have been integrated with GIS include spatial autocorrelation analysis, spatial regression techniques, exploratory spatial data analysis, and spatial point data analysis techniques, etc. From the deductive approach, a model-based approach will lead GIS users more quickly to working in a theoretical context. The researchers in this category favor the integration of more deductive models (for examples, dynamic modeling, Garin-Lowry modeling based on entroy maximization, micro-macro simulation modeling, and input-ouput/shift-share analysis) which support your hypotheses. This approach argues that only a theoretically informed GIS can be theoretically informative. I hope you find it to be informative. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1366) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 19, 1998 01:07 PM Author: Molly Brown Cisse (mcisse@glue.umd.edu) Subject: People and Pixels: Oil and Water? As a newcomer to the "GIS Environment", it is hard for me not to be overwhelmed the vocabulary you all seem to be using. However, from a the point of view of someone interested in human-environment relationships, I would like to comment on the white paper's positive review of using GIS for social and cultural information. As a review of research needs, I was surprised to see no mention of the lack of comparability of the "social, cultural and economic" data with the "environmental" data. Especially in the developing world, there seems to be nearly a complete lack of data on social systems that is on the same scale as the environmental data that exists. In the developed world, there is real difficulty in representing social phenomenon (like attitudes about housing) in a GIS database, and then relating changes (like in vegetative cover) seen over time (see "People and Pixels" 1998). Although we have large datasets for the US and Europe, we have a long way to go in having real confidence in social indicators that can be mapped on a large scale in areas of the world outside of the US and Europe. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1266) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 20, 1998 12:20 PM Author: Wilmot Greene (mot@uga.edu) Subject: Paper by Parsons Hi Molly, I know what you mean about the jargon, it can be overwhelming. There is a paper by Ed Parsons on visualization techniques for qualitative data. Basically, this article describes ways to integrate information that provides a "sense of place" into a GIS. This type of info. Can be presented by anything from photos to virtual reality tours according to Parsons. The article is on the web at http://www.odyssey.maine.edu:80/gisweb/spatdb/egis/eg94046.html I am personally interested by these types of new visualization techniques. The whole concept of spatial analysis should be founded on the ability to get a feel for the space in question. Sometimes all the number crunching can take center stage, but a more emotional sense of space (quirky as it sounds ) can be of great value to any spatial scientist. Where can I find "people and Pixels"? sounds cool. Wilmot Attachments: http://www.odyssey.maine.edu:80/gisweb/spatdb/egis/eg94046.html (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1282) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 06:52 AM Author: Molly Brown Cisse (mcisse@glue.umd.edu) Subject: that reference... Hi Wilmot: Here is that reference: Liverman, D., Moran, E.F., Rindfuss, R.R., Stern P.C. (1998) People and Pixels: Linking Remote Sensing and Social Science. National Academy Press: Washington DC. This is exactly what my research is on - using remote sensing time series analysis to tell what is happening in the environment, and then asking people about how the changes in the environment have affected them and their lifestyles. Finding ways to represent both of these in a GIS in order to do spatial analysis is the challenge. Thanks for the article - it provides some interesting insights. Molly (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1299) ------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 09:05 AM Author: Bill Moseley (wmoseley@uga.edu) Subject: Analysis of human-environment relationships Molly: Great to see that someone else is interested in human-environment interactions. My research will also focus on how people are dealing with environmental change (and given your surname, I suspect we may be looking at the same country, i.e. Mali). I will definitely check out the reference for the paper you gave. I agree that the white paper does not adequately deal with the problems associated with the analysis of environmental and social data. For example, how do we account for differences in the accuracy of data that has been remotely sensed and data that has been collected via interviews? Cheers! (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1354) -------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*] [Delete*] Date: September 22, 1998 09:45 AM Author: art getis (arthur.getis@sdsu.edu) Bill, There are several other white papers in the UCGIS series that address the issues you raise. Data quality is one of the chief concerns of the UCGIS. Clearly, there is just so much about data that can be discussed in a white paper on spatial analysis. Cheers! Art Getis (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1355) -------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*] [Delete*] Date: September 23, 1998 08:48 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: Human environmental relationships and profit Bill, There is another aspect to the problem you referring to (we have good coverage of N. America and Europe, but coverages of other places are lacking [ie. Africa]). Ask yourself why this is the case, then ask yourself to what extent GIS technology is profit driven. Of course, in the long-term and at the societal scale of things, GIS projects on vegetation change and human-environmental relationships in Africa might be profitable (if a GIS eventually leads to crop/grain self sifficiency, the North would save money with regard to donated grain abd emergency aid projects). But GIS technology, as it is, is woven into the current neoclassical (profit driven) way of things. The technology is expensive and is sold to the highest bidder. The market dictates that costs (for GIS technology) be recovered in the short-term. I don't see much potential for the kind of 'humanitarian' applications of GIS you're proposing. This is unfortunate and needs to be reversed. The above is an ethical consideration and probably belongs in the "GIS and Society" paper...but I can't wait. I guess the difficulty I'm having (in waiting) is this: in less developed nations the spatial analysis cannot take place...because the data is incomplete....and the data is incomplete because collecting data in such places is not profitable - Lynne (our faculty rep) is right - all of these topics are connected, one to the others. R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1385) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 07:24 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Spatial data acquisition and integration , and Scale I think what you commented is more related to other white papers under the title of spatial data acquisition and integration, GIS and society, and Scale. Anyway, it is not surprising that the social, cultural and economic data are uncompatible with the environmental data when doing research especially in the developing or the developed world. We need to collect all kinds of data from multi-source, multi-scale, and multi-resolution. That's why spatial data acquisition and integration techniqeus are so important in these days. I think the only solution to the problems which you suggested is to integrate GIS, remote sensing (RS) with socioeconomic data. As far as I know, several articles from the professional GIS and RS related journals deal with this kind of topic. Finally, it is true that there is real difficulty in representing human behavior and perception variables in a GIS database. However, it does not mean that there is no solution to the problem. Let me give you some references about this stuff. 1. Aitken, S.C. and Prosser, R., 1990, "Residents's spatial knowledge of neighborhood continuity and form", Geographical Analysis, Vol. 22, pp.301-326. 2. Aitken et al., 1993, "Neighborhood integrity and residents's familiarity: Using geographic information system to investigate place identity", Tijdschrift voor Economische en Sociale Geografie, Vol. 84, pp.2-12. 3. Can, S., 1992, "Residential quality assessment: Alternatives using GIS", Annals of Regional Science, Vol. 26, pp. 313-321. I wish you good luck. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1367) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 23, 1998 06:43 AM Author: Molly Brown Cisse (mcisse@glue.umd.edu) Subject: Thanks! Thanks for the references! I appreciate your comments. I look forward to the challenge of representing attitudes and experiences in a spatial format. Molly (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1379) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 23, 1998 08:17 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: PCV? Hey Molly, you write like a person who has had some international experience. Are you a Returned Peace Corps Volunteer? R. Ward RPCV 1993-96 Tanzania East Africa (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1384) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 30, 1998 08:01 AM Author: Molly Brown Cisse (mcisse@glue.umd.edu) Subject: Yes, of course... Hi Ron, Yes, of course I am an RPCV! I was in Senegal from 1992-1995 and sorely miss it, I must say. I am now in a doctoral program at the University of Maryland in the Geography Department. I am studying the affect of increasing climatic variability on local communities in the north and the south of Senegal. I am planning to use coarse resolution AVHRR data to analyze variability over an 18 year time frame and fine resolution data to create maps of the two areas. However, I dispair of putting any socio-economic information on top of these as there are such problems with comparability and accuracy. It is far to easy to gloss over all the multiple sampling and data collection problems the governments in Africa experience and pass the data off as equivilent to data obtained about the US. To the "uninitiated" US reader, it would seem as though I have created a wonderful, explanitory tool with a GIS, where all I have done is created the illusion of completeness. For this reason, I may not use any socio-economic data for anything other than a simple table listing population, birth, death and fertility rates in the 80s and 90s! Pretty low tech, but that is what the data deserves! Spatial analysis is interesting, but very frustrating in the data-poor areas of the world. Molly (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1600) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 20, 1998 08:16 AM Author: Jay Raiford (jraifor@lsu.edu) Subject: Greetings Hello out there in virtual land! Just a short intro, my undergraduate degree is in Forestry so my slant on GIS will come from that direction. First I would like to comment on Erik's thread about stats courses. Being a forester I have taken quite a few stats classes. Coming into geography I was amazed at the different way of looking at the data. We were more interested in numbers not locations. I beleive that mapping science students, especially me, would take more away from the spatial statistics by applying it, as Erik suggests, within the GIS. Being able to visualize real world data in an environment that we are going to use daily seems important. Statistics, like calculus, mean so much more to me when represented visually. My thesis is going to be on wildfire in the intermountain west. There will be a lot of spatial analysis involved. I hope to come up with new ways of using spatial analysis in predicting the spread. Hopefully some of our discussions can include real life scenarios to tie all of the theory together. Reading and discussing the white papers is a start, but I hope we are able to apply some of what we are doing (thesis, projects, research) to the discussion. This will help me personally in the understanding process. Have a great week! ..................................comments are always welcome (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1277) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 20, 1998 11:28 AM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Re:Greetings I think Jay made a good point about direction in the discussion. I agree with Jay. Firstly we need to review the white paper. In the next step we need to bring what we are doing to discussion. I believe it will drive us a constructive discussion. By the way, I'm interesting in what you are doing. What kind of methods (numeric simulation model, cellular automata, and GRID module in Arc/Info, etc) will you employ to predict the spread of the wildfire in the intermoutain west? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1281) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 05:44 AM Author: Jay Raiford (jraifor@lsu.edu) Subject: Re: Greetings I am hoping for the Montana Dept of State Lands and the Flathead National Forest to use my results. Because of this I will need to have the final product in ArcView($$). So far I have concentrated on the initial attack portion of my project. This will include the use of Spatial Analyst for locating available drawing and dipping water resources available to fight the fire and Network Analyst for routing. Homes in the study area will be risk rated for fire and I will need to know there spatial proximity to water also. After the vegetation is divided into the 13 categories, the USFS uses, and other criteria like topology, aspect, fuel moisture, weather etc. or put in the many formulas for spread will need to be incorporated for predictions. Unsure of the possibilities of being able to incorporate real-time weather. I'm trying to get my hands on several existing packages now to see how they handled some of my questions. Have only found one, so far, that incorporates GIS. I have not done anything with GIS that, for example, would make a fire grow, for testing scenarios, but look forward to the challenge. I understand that the USFS in Georgia is using Pacific Merdian Resources FIRE! model. Wish I was in Atlanta, would like to see it first hand. BTW, sounds like you have an excellent GIS/Forestry conference coming the end of this month. Financially unable to attend but it sounds very informative. As always suggestions and comments are welcome. Hope these type of comments are what the professors had in mind. If not we can continue communicating directly. Somehow I feel we will know soon. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1297) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 20, 1998 05:46 PM Author: Kurt L. Johnson (Kurtljohnson@worldnet.att.net) Subject: Spatial Analysis in a GIS Environment Hi Everyone, First of all I would like to thank all of you for your many thoughts and comments. I found them to be very informative and stimulating. I know this type of open discussion will be very beneficial for me, I hope it will be the same for you too. Much of this white paper is concerned with how, we as scientists, can utilize the increasingly vast amounts of spatial data available to produce quality answers. As many of you have said, despite the technological advancements made in computer design, our ability to analyze data at a rate equal to the rate at which we collect data is severly behind. This I believe is partly a problem with our training and exposure to different kinds of data. Erik's comment suggesting the use of more GIS in our statistical training is a significant point. After all, our interest as scientists is to present data intelligently so that it is informative and representatively defensible. To achieve this goal we must do more to integrate GIS in statistical analysis. If we would utilize more visualization, we would become more proficient in our use and comprehension of the interactions between the many diverse data sets that exist. This is not to mention the increase in the comfortability factor that I think all of us would appreciate. We will make mistakes but I have to say, I need to make mistakes to understand the full extent of what is going on. On another issue pertaining to spatial analysis, I think the initial analytical approach should be of an exploratory nature. This would serve to allow the analytical process to proceed ojectively, and through visualization help to identify subtle patterns or trends unique to those data. I also think it would help to identify the advantages and disadvantages concerning the relationships between different spatial units. The resolvement and understanding of these relationships are inturn, important when undertaking a confirmatory approach. Ideas and comments are welcomed. Thanks: kurt at LSU (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1293) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 09:43 AM Author: Esra Ozdenerol (kalem1@hotmail.com) Subject: More research building towards a better established framework GIS is the most promising technology to demonstrate any analysis with spatial structure. It allows enhancement of existing data collection systems and identifies data gaps in existed databases. Advances in computer technology, geographical data systems, and statistical methodology have meant that a variety of Spatial Analytical studies are now feasible with GIS. However, the type of study undertaken and the methods required depend on the question to be answered. For some kinds of inquiry, the geographical, statistical methodologies are embedded within a standard and accepted framework. But there are some inquiries that an established framework does not exist. At this point, the scientific community needs to advance its research. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1308) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 09:51 AM Author: Esra Ozdenerol (kalem1@hotmail.com) Subject: Summary of Digital Earth and 5 research priorities Esra Ozdenerol Geography 7973-Advanced GIS Seminar September 13, 1998 UCGIS stands for University Consortium for Geographic Information Science. In Columbus, Ohio, delegates of UCGIS gathered and came up with 10 research priorities for Geographic Information Science. This two page summary includes 5 research priorities which will not be covered by this seminar. Those are: Interoperability, Spatial Information Infrastructure, Uncertainty, Spatial Data Acquisitions & Integration, Cognitions. "The Digital Earth: Understanding our planet in the 21st Century" will also be summarized in these two pages.. Interoperability: One can not see any GIS without transferring data from one system to another; to access one system's data from another; to control one system with the commands defined for another, or to take experience accumulated with one system and apply it. In other words, GIS can not survive without exchange of information between different systems. A key component of any interoperable environment is a shared system for describing data. For systems to be interoperable there must be a consistent set of interpretations for information. One system must be capable of understanding the meaning of another system's data. Research is needed to develop standardized languages for describing operations and build tools that are able to find commonalities between data from different systems and agencies. Methods need to be developed that are capable of extracting and updating essential metadata automatically. Spatial Information Infrastructure: Through development of geographic data, there is a lack of knowledge and experience of the complex policy-related issues that arise from the creation, compilation, exchange, and archiving of large geographic data sets. The ownership of digital geographic data, protection of privacy, access rights to geographic data held by governments, and information liability issues require greater clarity. Research is needed to identify optimal government information policies and practices for promoting a robust spatial information infrastructure. Intellectual property rights, information privacy, and liability are the basic policy issues. Public and private roles in information creation through partnerships and cooperative agreements need to be examined. Being an unusual commodity of great value, geographic data and its cost recovery, pricing and marketing are of central importance, too. Developing the technical and institutional means to support creation and contribution of local knowledge presents a challenge to technologists and decisionmakers. Uncertainity: Uncertainity information can be conceived as a map depicting varying degrees of uncertainity associated with each of the features represented in the data set. The typological attributes (describing the type of a geographic feature), the locational attributes, and the spatial dependence (the spatial relationship with other features) are the three attributes that are subject to uncertainity. Forest environment can be a good example. The species and type of the forest is typological attribute. The forest's location and its own extend is locational attribute. The neighboring landscape near the forest would be Spatial Dependence. Since stored information is only an approximation to reality; they may also change overtime. This makes geographic data very complex and difficult to manage. Uncertainity exists from data collection to data representation, data analyses and final results. It is a measure of the difference between the data and the meaning attached to the data by the current user. New strategies need to be implemented for reducing, quantifying, tracking, and reporting uncertainity in GIS, geographic data collection and generation. New methods need to be tested to manage uncertainity in GIS analyses and geographic data. Spatial Data Acquisition & Integration: By the advance of new technologies, GIS, Remote Sensing and GPS, Computer Technology, Geographic Information Science is emerging rapidly surrounding the topics of capture, storage, analysis, interpretation and communication of geographic information. In order to develop better tools for data integration, high quality research is needed. Techniques that are capable of automatic registration of geographic data sets , based on recognition of common features, and adjustments to both geometric positions and feature types need to be developed. Effective research on integration will require the collaboration of many sciences with common interests and motivations, including image processing, pattern recognition, robotics, computer science, geodetic science, and photogrammetry. There will always be continued research into better tools for spatial data acquisition. Ground-based data acquisition systems will be advanced. Sophisticated computer algorithms for directing ground-based sampling, recognizing patterns, and analyzing data directly in the field will be employed. Cognition of Geographic Information: Cognitive research will lead to improved systems that take advantage of an understanding of human geographic perception and expertise. There are inexperienced and disadvantaged users that are not accessible to geographic information technologies. Cognitive research can fill this gap. This kind of research holds great promise for improving geographic education at all levels, by addressing general concerns about poor levels of geographic knowledge in society and low levels of awareness of such critical issues as global environmental change. Internet allows users to search for digital geographic data over the network. Future research needs to focus on providing a user interface that successfully reproduces all of the map library's functions. The Digital Earth: Understanding our planet in the 21st Century We are in the era of new technological innovations. There is an insatiable hunger for knowledge but still a great deal of data remains unused. Al Gore believes that it is time to think about a "Digital Earth" which is a three-dimensional representation of the planet, into which we can embed vast quantities of geo-referenced data. This is such an imaginative idea that can be a good education source for young generations. At local museums, whatever a young child is interested in exploring about an area of the planet, she can request information on land cover, distribution of plant and animal species or visualize the environmental information through GLOBE project. Using a data glove, she can zoom in, using higher levels of resolution, to see continents, regions, countries, cities, and individual houses, trees, and other natural and man-made objects. She can also travel through time. She can move backward in time to learn about history and send some of this information to her personal e-mail address to study later. Not only for education purposes, "Digital Earth" might become a digital marketplace for private companies selling information. It has the potential to be a laboratory which allows the complex interaction between humanity and our environment. There are technologies that are required to build this fantastic "Digital Earth" idea. Computational Science helps to overcome the limitations of both experimental and theoretical science. Modeling and simulation will give us new insights into the data that is collected about our planet. Mass Storage issue is getting improved by NASAs Missions to Planet Earth program. Satellite Imagery will provide 1-meter resolution imagery which will give incredibly accurate results. Broadband networks is high-speed networks that digital Earth need to be connected. The Digital Earth will need some level of interoperability and meta data quality will be improved by voluntary standards for metadata. Digital Earth has to be perceived as a great opportunity which has both societal and commercial benefits in areas such as education, decision-making for a sustainable future, land-use planning, agricultural, and crisis management. Rather than being maintained by a single organization, it would be composed of both publically available information and commercial products and services from thousands of different organizations. As a conclusion, "Digital Earth" is not a project that will happen overnight. A lot of parties and groups like university researchers, science museums, local schools should work together to enrich this project. Experts in government, industry, academia, and non-profit organizations need to be challenged to develop a strategy for realizing this vision. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1310) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 10:11 AM Author: Guangxiang Cheng (gcheng1@tiger.lsu.edu) Subject: Discuss of digital earth and other five research priorities Geography 7973 - Advanced GIS Seminar Guangxiang Cheng September, 10, 1998 This report is from reading the research priorities of white paper of UCGIS internet homepage and Mr. Gore's speaking. The Digital Earth The Digital Earth is a project proposed by the U.S. government. This will be a geographic information system integrating may geographic data resource. The system will have the high-resolution planet image, digital maps and economic, social information. Also the Digital Earth will be used in many industries, government, organization to gain benefit. The first stage of Digital Earth is to integrate the different data sources. The next stage is to develop a map to cover the whole world at one meter resolution. The final goal is put full data rang and history on the earth. Cognition of Geographic Information The research of cognition of geographic information is to study the relationship between human being and the geographic information knowledge. A better knowledge of human being's perception, memory, reasoning, and communication with geographic information knowledge can help the researchers on the GIS priority research and GIS' application. Six categories in this field are considered as high priority. Those are: 1. the limitation of inconsistencies in current model caused by human cognition and how to deal with; 2. how to design and implement intelligent transportation systems; 3. how can GIS incorporate natural language; 4. how can spatial metaphors be used to represent and manipulate information; 5. How can GIS be used to represent and communicate in new methods; 6. how to apply the new virtual environment technologies. Interoperability of Geographic Information Interoperability of Geographic Information is to study the interchange of spatial data from incompatible systems. The object is to develop a formal knowledge-based integrated languages to make the spatial data from different domain systems communicable. This research field has short-term and long-term goals. The short-term goal is to provide a more complete formal specification of the semanics in GIS and the use in intercommunication among systems. The attempts include identifying the domain-specific questions, modifications to models and the role of scale, data statistical support and model assumptions to answer these questions. The long-term goal is to develop languages, semantic theory and geographic knowledge representation to support GIS and construct a digital libraries of global information. Short term results will be used in the long-term projects. The future of the Spatial Information Infrastructure A spatial information infrastructure is composed of effective policies, strategies and organizational arrangements. The spatial information infrastructure can be used in government, industry, academic an public sectors. The infrastructure can help those sectors to have a better understanding among geographic-related alternatives. Three tenets are the basis for the spatial information infrastructure - technology and policy, institutions and traditions; technical facts; and information policy issues at all level. Several factors have influence on the research. They are the stimulation of economic growth in geographic information industry, the strengthening of institutional capacity and the promotion of democratic processes which is to let public easier to access the government information. The priority of this research field are in four areas - 1. information policy, 2. access to government spatial information, 3. economics of information, and 4. local generation and integration of spatial information. Uncertainty in Geographic Data and GIS-Based Analyses This research priority is to understand the uncertainty in geographic data and how to deal with it is data analyses. Methods must be developed to identify, quantify, track, reduce and report the uncertainty in geographic data and GIS-based analyses which is vital to many fields. Understanding, quantifying and visualizing the uncertainty will save tremendous amount of energy and money during data collection and help to develop a healthy geographic information science as a field. The research will be done in three stages: short, medium and long terms. In short term, test techniques to measure the uncertainty will be developed. The effects of uncertainty will be presented. In medium term, the uncertainty from incompatibilities of spatial data, various stages of GIS life cycle will be studied. In long term, a complex dynamic representation will be developed to minimize the uncertainty in spatial data. The interaction between forms of uncertainty generated at each stage in the life cycle of GIS data will be clearly understood. Detailed discussion Spatial Data Acquisition and Integration The spatial data collection methods has been improved tremendously qualitively and quantitively. However, technology of spatial data collection has surpassed the technology of integrating the captured data. This then lead to the waste of data resource. The development of the data integration will reduce or eliminate some costs related with data collection and then get a better database quality. This technology is a kind of interdisciplinary which includes cartography, computer science, photogrammetry, geodesy, mathematics, remote sensing, statistics, modeling, geography, and various physical, social, and behavioral sciences with spatial analysis applications. My background is in civil engineering and transportation planning. The better data collection with effective data integration can be applied in both engineering design and transportation planning. Using other data set, if it is accurate enough, can certainly decrease and even eliminate the cost of data collection. Using the tool of data integration in the software plus the availability of the spatial data, the job can be done with much less cost. For instance, transportation planning is to predict a region's future transportation situation. The map of the planning area is necessary for transportation planner. If the TIGER database (Topologically Integrated Geographic Encoding and Referencing) can have enough accuracy, the transportation planner can certainly use the regional spatial database from TIGER. This can save lots of money. If a geographic information system can be developed more detailedly both in data collection and integration, for instance, 0.01 meter accuracy horizontally and vertically, provided the data collection and integration methods have been developed very advanced, accurate and efficient, the database can be used by civil engineer in roadway design. This can save huge time in survey for civil engineering consulting company. Certainly this databases would be so valuable that can be used in many industries and government departments. This is just my imagination. This system should even be much more accurate than the Digital Earth which will have a resolution of 1 meter and was presented by Mr. Al Gore. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1311) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 10:23 AM Author: art getis (arthur.getis@sdsu.edu) Subject: Digital Earth Now that things are rolling along, I thought I would add a recent statement for your review (see below) that Harvey Miller and I drafted for UCGIS with regard to Digital Earth. Some of the ideas in it will be included in our 1999 white paper. What do you think? Art Getis Statement by the Spatial Analysis and Modeling in a GIS Environment leadership of the UCGIS with regard to Vice President Al Gore's speech on The Digital Earth given on January 31, 1998. Vice President Gore's comments on the "Digital Earth" initiative recognize the importance of spatial analysis in helping society to better understand the Earth. Knowledge of how physical and socioeconomic geographic processes evolve is critical to our understanding of the Earth as an interrelated physical, social and economic system. Applying this knowledge can aid society in resolving increasingly difficult problems involving the efficient and equitable distribution of resources and the impact of resource use on the Earth's environments. However, geographic phenomena are interrelated in complex ways. Failure to recognize and capture the complexities of spatial relationships will result in inappropriate conclusions and the misapplication of scientific research in solving the world's environmental and resource distribution problems. Geographic phenomena are related in subtle and multidimensional ways, potentially involving factors such as distance, proximity, connectivity and direction, all intertwined and mitigated by the geographical context of the phenomena. Also critical is spatial heterogeneity, i.e., changes in the observed characteristics of a phenomenon with changes in location. A strong knowledge of spatial theory and analysis is required to avoid reductionism and simplification of these subtle geographical relationships when handling, processing and communicating georeferenced information. The spatial analysis community within the University Consortium for Geographic Information Science is attempting to develop theories, models, statistical methodologies and computational techniques that capture and communicate the full spectrum of information available from a multi-scale digital representation of the Earth. A major thrust of this research is geocomputation techniques that support processing terabytes of geographic information quickly and effectively within appropriate modeling and analytical frameworks. Despite their power, computational techniques such as massively parallel processing, numerical simulations, knowledge-based databases, intelligent agents and visualization cannot replace missing or ignored information on geographic relationships. Spatial analysis provides the framework for ensuring the appropriate application of computational science within the Digital Earth initiative. Visualization provides an excellent case-in-point. A naïve approach to visualizing geographical relationships is to use geographic proximity as the major organizing factor. This creates the digital analog to the centuries-old paper map. While proximity often has a profound effect, geographic relationships can also involve varying degrees of spatial and temporal lags operating at varying scales as well as hierarchical and indirect effects. While traditional mapping has difficulty in capturing these effects in a satisfying manner, the emerging computational environments are potentially liberating and even revolutionary. To realize this potential, computational scientists and digital cartographers must work closely with spatial analysts to formulate entirely new methods to visualize geographic data and relationships that expand greatly on the traditional paper map. Neglecting spatial dependencies and spatial heterogeneity can result in inappropriate scientific conclusions and misdirected policy. Resolving key issues for the twenty first century such as crime prevention, preserving biodiversity, recognizing and mitigating human-induced climate changes and improving agricultural productivity require recognizing potential spillovers and unintended "boomerang" effects that can impede if not negate progress. These effects often result from not recognizing or ignoring subtle and complex spatial relationships among the phenomena of interest. Spatial analysts, working closely with other geographic information scientists, can provide an understanding of the Earth that minimizes the unintended consequences of applied science. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1312) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 07:25 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: comments on the proposed 1999 white paper on 'digital earth.' This paper brings up important points regarding the use and abuse of spatial data. I'd like to comment on the idea of reductionism. Traditional scientific philosophy is reductionist (trying to reduce complex relationships to basal causes and effects). In many areas of science researchers are trying to depart from reductionist thinks, and instead, are trying to develop new ideas that encompass the true complexity of what goes on in the real world. In biogeography, successional theories developed by Odum, which were an attempt to make sense of spatial dynamics in forested landscapes by theorizing an 'equalibrium', have given way to new ideas. these ideas are more inclusive of the complex, constantly changing nature of a forest. Now, thinking in terms of 'patchwork mosaic' systems, in which all part of the forest are in a different place along the early succession to old growth continium, has become the norm. These are important ideas with regard to GIS and the need to develop methods of analyses that can handle large data sets and analyze relationships between large numbers of variables. These ideas are also relavent to spatial data cognition, for as Mot Greene has said, "visualization is key." Complex analytical methods are all well and good, but without representational techniques capable of displaying complex information in a manner by which people can easily understand, spatial analysis and representation will become the playground of the redutionist elite. Good paper! R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1348) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 01:55 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Comments on the Proposed 1999 White Paper on The Digital Earth I think this paper documents well the importance of spatial analysis and its two future research agenda(such as computational intelligence-based spatial analysis and visualization) in the context of the Digital Earth initiative. The importance of spatial analysis within the Digital Earth initiative is that a strong knowledge of spatial theory and analysis is required to avoid drawing out inappropriate scientific conclusions and misdirected policy in data-rich and computer-rich environment. The paper also addresses two major research directions in the context of the Digital Earth initiative. One focuses on developing techniques of computational intelligence (CI)-based spatial analysis which provides a basis for improving the spatial data analysis techniques and models to meet the large-scale data processing needs of the emerging new era of data-driven exploratory searchs for patterns and relationships in the context of an analysis process increasingly driven by the availability of very large quantities of spatial data. In this case, as the paper indicated, "spatial analysis provides the conceptual framework for ensuring the appropriate application of computational science within the Digital Earth initiative." Another focuses on the development of advanced visualization techniques which can handle various statistical effects in visualizing geographic data and relationships. Visualization techniques in GIS are very useful to especially exploratory spatial data analysis. I think this paper is an extension of the former white paper. The former paper already described the basic concepts (about geocomputational techniques and visualization) which are mentioned in the proposed paper. It seems not to be new to me, but it is well-documented to this topic. According to my understanding, the Digital Earth initiative is to apply spatial information technologies and interoperability toward overcoming a wide range of challenges and problems. If so, how can the interoperability work in terms of spatial analysis? Is spatial analysis enough as a conceptual framework of how physical and socioeconomic geographic processes evolove? What about social theory? Comments and opinions? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1362) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 25, 1998 09:29 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: social theory and GIS Hey Jun, for background information on GIS and social theory (social theory and GIS interactions) see 'Cartography and Goegraphic Information Systems,' Vol. 22, No. 1, 1995 - specifically, read the introduction article by Eric Sheppard (the social theorist from the University of Minnestoa, not the Erik Shepperd in our GIS class)and the GIS and social theory article by Roger Miller (also from the U of MN). You asked the question, "what about social theory." In these articles you'll find more specific questions along that line. In general these articles outline how GIS are influenced by society with regard to rpofit motives and the commdification of information (GIS 'should' generate profits), and how society(ies) are influenced by the types of information available in GIS representations. Good stuff! R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1481) ------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 25, 1998 10:07 AM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Re:Social Theory and GIS Thank Ron for giving me information. I think I have them. You're right. I belive Dr. Sheppard's article called "GIS and Society" provides us with a good reference for next discussion in the session on GIS and Society, and Dr. Miller's arcticle called "Beyon method, beyond ethics: Integrating Social Theory into GIS and GIS into Social Theory(1992)" argues that GIS representation of reality constructed both socially and culturally in the Cartesian space is too limited to be adequate (called ontological inadequacy in GIS). What I was saying is that an integrated approach of spatial analysis and social theory would be good enough to explain how physical, socio-economic, and cultural processes evolve. Anyway, thanks, Ron. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1483) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 10:38 AM Author: Caiming Shen (cmshen@hotmail.com) Subject: Background After I read the white paper, I did not found what the spatial analysis can do at present. My meaning is that we need a background which tell us what kinds of statistical methods are being used in current GIS software, what kinds of problems those methods can be used to solve, how great data sets they can handle, whether those methods are appropriate for handle large spatial and spatial-temporal data sets, how far the spatial analysis in GIS has gone, and so on. I think it is important, especially to newcomers like me. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1314) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 10:45 AM Author: Guangxiang Cheng (gcheng1@tiger.lsu.edu) Subject: comments on spatial analysis Geography 7973 - Advanced GIS Seminar Guangxiang Cheng September, 20, 1998 This research priority- spatial analysis in a GIS environment clearly stated the objective, benefits and priority areas for research. More concentration will be on dealing with geographic data which are related with other variables. Much more techniques - mainly on computer science and geo-statistics will be applied to this area. New model will be devised to explore this area. This paper stated the importance to National Research Needs. A large range of social concerns has been introduced. Among those, the GIS technique in traffic management and land use planning is also introduced briefly. Transportation planning is one of my interest. GIS can have tremendous help in this area. One software package - TransCAD is based on GIS technique. This package contains the spatial analysis methods plus the transportation models and has been applied in many transportation agencies. The clearly stated importance gives research directions. This white paper also discussed the priority areas for research. The priority researches in spatial analysis also mentioned other research priorities such as scale, data integration and uncertainty. GIS researchers should have the knowledge of the whole area of GIS research priorities. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1315) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 11:22 AM Author: Nina Lam (ganlam@lsu.edu) Subject: metadata, spatial data analysis, and others Hi all: I am delighted to see the active postings. Hmmm, seems like we are running an SEC football conference here, LSU vs. Georgia vs. others!! I encourage you all continue in actively participating in the discussion. I like to clarify three points: First, since we have semester system and started on Aug. 24. I asked my students to work on the other five UCGIS research priorities and digital earth for the last three weeks and asked them to post their writeups. That's why you will probably see a flood of postings. I also asked them to join the spatial analysis discussion last week and this week and posted the writeups by next Monday (Sept. 28). Is that what we shoud do for the 5 designated topics, Dawn and Art? Second, a brief comment on the two terms spatial analysis and spatial data analysis. To me, they are the same. Spatial data analysis is the basic of spatial analysis, and I would not consider a piece of research as spatial analysis if there are no data involved. Third, about meta-data that Ron mentioned. Yes, I think the prime purpose of metadata is for easy search. However, the existing metadata community (FGDC) seems to have focused on the other aspects of data than their content. I am an advocate for including description or index of data content as part of metadata. Furthermore, I am advocating that we should include spatial indices. I am currently working on the use of fractal indices, spatial autocorrelation indices, and other landscape indices as a quick means for data mining, environmental assessment and monitoring. Nina (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1316) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 07:31 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: clarification request Nina, Thanks for clarifying some of the questions I had about spatial analysis vs spatial data analysis, and my question about how metadata fits into all of this. Could you give me an example of how this metadata index you proposed could be used for data mining? I'm sitting here thinking about it, but I'm a little slow, and I usually need a working example before I get a point(s). R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1349) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 23, 1998 10:39 AM Author: Dawn Wright (dawn@dusk.geo.orst.edu) Subject: posting writeups > flood of postings. I also asked them to join the > spatial analysis discussion last week and this > week and posted the writeups by next Monday > (Sept. 28). Is that what we shoud do for the 5 > designated topics, Dawn and Art? Hi Nina, This sounds good to me. Forgive my brief "disappearance" from the seminar but I've been travelling on business since the 16th. I'll be back in my office on the 26th. I'm writing to you from Charleston, SC in the middle of an FGDC clearinghouse workshop (another interesting topic!). Better go and get back to paying attention to the presentation! Cheers, Dawn (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1407) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 12:54 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: FYI: Al Gore's Speech on the Digital Earth I'd just like to provide Vice President Al Gore's speech source on The Digital Earth for you who are interested in it and couldn't reach it. I hope you find it to be very informative. Attachments: ALGORE.HTML The Digital Earth: Understanding our planet in the 21st Century by Al Gore. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1321) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 21, 1998 07:19 PM Author: Jay Raiford (jraifor@lsu.edu) Subject: Digital Earth plus 5 topics not covered Dr. Lam requested us to post our first papers. Attachments: Summary - UCGIS.doc, Digital Earth.doc (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1332) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 12:24 PM Author: xiaojun yang (yang@uga.edu) Subject: SOFTWARE, the Key of Spatial Analysis? As a "cypervisitor", I have seen so many different perspectives on the spatial analysis that are posted in this discussion forum. Now that we have many good ideas, what is the next we should do? Argue in theory or go down to actual job? Do everybogy agree that SOFTWARE should be the key for spatial analysis? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1357) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 01:22 PM Author: Erik Shepard (shepard@uga.edu) Subject: Opinions and thoughts There has been such a flurry of postings and ideas, that I thought I would try to just throw out some of what I think. There are a couple of "camps" here that I see (please excuse my attempt to create some categories which are deliberately simple): - the software-ist category which allocates the creation of software tools to faciliate GIS based analysis - the algorithmist category which desires to create algorithms with which to do spatial analysis on new data types and models (closely related to software-ism but a little more theoretical in foundation) - the curriculum-ist category which seeks to integrate GIS and spatial analysis (either via spatial analysis lessons in GIS courses or via GIS familiarity in spatial analysis courses. Note that to do GIS in spatial analysis courses, it is necessary to have some software tools from the first category already built and functional). - the visualist category which advocates the improvement of visualization techniques to understand the results of analyses. This seems to me to also require the existence of some software to produce the results that need to be visualized in the first place - the scientific methodlogist category which seeks to understand the role of these spatial analysis tools within the context of the scientific method (a la exploratory versus confirmatory, etc) -- the applications category which wants to know what applications there are for this integrated spatial analysis / GIS (in other words, why do we care?) Personally, I like seeing all of these discussed and I advocate particularly the creation of software and algorithms coupled with an understanding of the theoretical basis of these spatial analyses together with some mechanism for viewing and understanding the results. While I also think that there is a definite need for understanding how this topic fits both into theory (the scientific method) and applications, I think that these topics are more tangential to the original discussion in the whitepaper of "just how do we integrate GIS and spatial analysis". Comments? (I am fully expecting to be flamed quite thoroughly for this monologue...) (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1358) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 01:53 PM Author: Bill Moseley (wmoseley@uga.edu) Subject: Question regarding distinct treatment of spatial data At the risk of sounding like a GIS outcast, why must spatial data be treated differently from other types of data? Isn't the vast majority of data somehow related to space. I know it probably takes an entire course to answer the above question, but I would be grateful for a brief reply. Many thanks! (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1360) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 24, 1998 12:53 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Re:Bill's Question I think Dr. Anselin in 1990 gave us a good answer to your qestion. Briefly, the point is that spatial effects (including dependence and heterogeneity) complicate any straightforward understanding of spatial data. For more detail, please refer to this article: Anselin, L. and Getis, A., 1992, "Spatial Statistical Analysis and GISs", The Annals of Regional Science, Vol. 26, pp. 19-33. Good luck! (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1457) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 22, 1998 11:14 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Progress in Spatial Analysis and Modeling in a GIS Environment I'd like to give you a review of progress in spatial analysis and modeling in a GIS environment. It will help you better understand the recent trends in GIS and Spatial analysis/modeling communities. The 1990s have witnessed the increasing involvement of quantitative geographers and regional scientists in GIS research. The recent GIS literature in spatial analysis and modeling can be grouped into three broad categories. First, there are growing efforts in the GIS community to integrate spatial analysis and modeling techniques developed in the 1950s and 1960s with the GIS of the 1980s and 1990s. Actually, the cross benefits of the close link between spatial analysis and GIS have formulated such an integration. The mutual benefits include spatial data handling and visualization capabilites from GIS and spatial statistical capabilities from (spatial) statistics. Four different paths toward such an integration have been developed, including the stand-alone module, loose coupling, close coupling, and full integration approaches. Second, there are increasing applications of GIS-based spatial analysis and modeling to geographical research. The integration of GIS with spatial analysis and modeling resulted in a wide variety of fruitful applications in geographic researches, ranging from simple spatial pattern analysis to urban crime studies. Third, there have existed a need to develop new styles of spatial analysis models and methods in order to cope with spatial data-rich and computer-rich environemts which include GIS, RS and Computational Intelligence (The Digital Earth initiative). There are currently few relevant methods and every opportunity to use existing knowledge and skills, and computational science should be used to develop new tools and procedures. The emphasis on new methods reflects the failure of old techniques, the lack of simple spatial analysis problems, and a poor or inadequate analysis prior knowledge, and a recognition that spatial analysis and modeling in a GIS environment is not at all easy or straitforward. The first and third categories are more methodologically oriented whereas the second category leans more toward specific applications. Although there have been an enthusiasm about such an integration between spatial analysis/modeling and GIS, no consensus on how such an integration should be made has reached. Spatial analysis community's position is that the technology should be led by theorectical and methodological developments in the field itself. However, GIS community's position is that the analysis should be led by the technological developments. Technology-led or analysis-led is now a buzzing word in both GIS and spatial analysis communities. Currently, spatial analysis and modeling appears to be a rich field with many linkages to urban and regional problems, and marketing, transportation, and natural resource problems. GIS- and RS-technologies are greatly increasing the need for spatial analysis. Any comment? (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1377) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 23, 1998 07:32 AM Author: Jay Raiford (jraifor@lsu.edu) Subject: Request I propose that we all back up one step here. These are very interesting comments from all. I do think however that if we knew a little about each other it would prove useful in interpreting responses. I think it would be nice if evreyone here, that has not already done so, go back to the welcome thread. Here we could give a little personal background on ourselves such as education level, degrees we have, interests we have, etc. What do you think? PS Maybe more people would post if we didn't seem so confontational in our responses. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1381) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 23, 1998 08:07 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: agreed - THIS IS IMPORTANT!!! I agree 100% with Jay! When replying to somebody's comments in a particular thread, it would be nice if we could use the welcome thread as a reference source to see the set of interests we are replying to (mine being low for example), the level of GIS expertese we are critisizing, and whether or not other people have shared interests we can use to create new threads on specific topics. THIS IS IMPORTANT!!! R. Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1383) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 23, 1998 12:05 PM Author: art getis (arthur.getis@sdsu.edu) Subject: Looking Toward the Future Threre are only a few days left for this type of discussion on this theme. It is time to take stock and then look forward. Let me pose a few questions that may be worth your time and attention to contemplate. 1. Is there a major (not minor) place for spatial analysis, in both its exploratory and confirmatory modes, in GIS? Should academic GIS be centered on using GIS as an analytical tool? 2. What types of analytical procedures show the most promise for the future (for examples: geostatistics, spatial econometrics, pattern analysis, studies of spatial autocorrelation, data reduction routines, classification schemes, etc.)? 3. What threads in our discussion seemed worth mentioning in the next white paper? What themes are provocative enough for us to give them further attention? 4. Has this discussion increased your interest in learning more about analytical techniques, or did it depress you (for any of a number of reasons)? If you decide to answer any of these in this seminar, please keep your answers reasonably short. As has been suggested: please say a word or two (short) about your own interests, so that we part this aspect of the seminar with a better idea of just who we are. Thanks for your participation. Best wishes, Art Getis Birch Chair of Geographical Studies Research on local statistics, and disease and ecological spatial distributions in a GIS environment. Co-Editor: Journal of Geographical Systems (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1433) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 23, 1998 12:59 PM Author: Jay Raiford (jraifor@lsu.edu) To me spatial analysis is GIS. We need, as students, to leave the university not only being able to understand GIS theories but to be shown how they are applied. It is also possible, as a new graduate student, to not be familiar with many of the topics you and others have brought up. As someone suggested the other day, I beleive if the white papers are going to be used to teach from they should give a much better understanding of today and the past, not just future challenges. I sometimes feel I have taken a course and skipped the prerequisites when I read the threads. My bio will be in the introduction section for easy future reference. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1434) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 24, 1998 02:20 PM Author: Erik Shepard (shepard@uga.edu) Subject: Re: Looking toward the future My general take on this issue is that I feel that there is a major place for spatial analysis in GIS. There are existing spatial analysis techniques which could be incorporated very easily into GIS to take advantage of the underlying data structure. If we already have the data in the computer, why not start using the computer to not only make maps with it but to really simplify things and let us begin to do some analyses with it. There are admittedly some rather large issues (particularly with regards to development of new techniques - both spatial and spatiotemporal - and techniques to deal with the large volumes of data we are beginning to aquire) but there is no reason the GIS can't be a number cruncher as well as a map maker. Who knows, when we start really using the GIS to analyze some of this stuff, we may get some of the coolest maps yet. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1460) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 24, 1998 08:38 PM Author: Byong-Woon Jun (bwjun@arches.uga.edu) Subject: Re: Looking Toward the Future I believe that there is a major place for spatial analysis, in both modes, in GIS. There have been a need to incorporate both explartory and confirmatory methods of spatial (data) analysis into GIS. In exploratory spatial data analysis (ESDA), we searches for structure and association while in confirmatory spatial data analysis (CSDA), we evaluates the evidence. GISs, being very data-rich and having excellent visualization capabilities would seem prime targets into which to embed ESDA routines. However, I think the major place for spatial analysis, in both modes, in GIS inlcudes measuring spatial association or spatial autocorrelation, and spatial regression model. For example, regressing one variable on several others might be cosidered confirmatory, whereas mapping the residuals from that regression in order to uncover relationships not already in the model might be considered exploratory. The view was held that it would be of considerable benefit to the GIS community if a good regression package able to fit spatial regression models were to be coupled with a working GIS. Just try it and do it. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1477) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: September 25, 1998 09:54 AM Author: Ronald William Ward (ronward@arches.uga.edu) Subject: re: looking toward the future In reply to the questions posed by Dr. Getis: 1. Is there a major (not minor) place for spatial analysis, in both its exploratory and confirmatory modes, in GIS? Should academic GIS be centered on using GIS as an analytical tool? After talking over the topic with our faculty rep (Lynne Usery), I'm leaning toward more research on exploratory methods as well as the usual confirmatory routes. With the amount of data available to GIS expanding exponentially as time goes on, there is a real need to develop better methods of making use of data which are costly to compile in the first place. The merits of confirmatory methods are, I believe, well documented. 2. What types of analytical procedures show the most promise for the future (for examples: geostatistics, spatial econometrics, pattern analysis, studies of spatial autocorrelation, data reduction routines, classification schemes, etc.)? I have to answer this question by asking another question: by most promise do we mean for societal problem solving, for demonstrating the utility of GIS in an economic sense, or both. GIS as potential for social good and as having economic applications are, of course, related - but what exactly do we mean by 'promise.' 3. What threads in our discussion seemed worth mentioning in the next white paper? What themes are provocative enough for us to give them further attention? Clarify differences between spatial analysis and spatial data analysis, a few participants do not believe there are any differences at all. A provocative theme might include expanding on, by use of examples, the need for exploratory research methods. 4. Has this discussion increased your interest in learning more about analytical techniques, or did it depress you (for any of a number of reasons)? Both. It's depressing with regard to the statistical inaccuracies brought up by a few participants, yet exciting with regard to the ways in which we all agree that new methodologies will help us handle space and space-time problems in ways we are only now beginning to realize/conceptualize. cheers, Ron Ward (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1482) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 10:01 AM Author: Mihye Bark (gabark@unix1.sncc.lsu.edu) Subject: General Opinion:Looking Toward the Future Spatial analysis is a very important and broad topic in a GIS environment. My general observation on spatial analysis of virtual seminar is following (It might be not only for this topic but also for every topic): We have many students with different background in the seminar. For our understanding and participation, we must have exact prerequisites for this class or some section or references which give us more chances to have our fundamental and deep understanding about theories and methods. If not so, only a few persons might participate actively in this class. Moreover, this seminar might turn to a general discussion of news groups like geographic information systems, or geostatistics. No one might attend this discussion. We have limitation to review references and discuss specific topics related those (in spatial and temporal aspects). For example, we can get and read articles very easily. In case of books, it is not easy to get books and even though we have books, we have so many students to read the books. If possible, I hope to have the systematic, detailed, or updated reference guide to get, read, and study appropriate theories, methods, and applications. Also, I feel that we occasionally need feedback from students to professors or vice versa for past topics within this semester without hurting the scheduled time and topics. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1676) ------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 11:07 AM Author: Doug Albert (dalbert.unix1.sncc.lsu.edu) Subject: U.S. at the cutting edge? I think to the best of my knowledge no one questioned the line "for the United States to remain on the cutting edge of GIS technology" (White papers 1998: 16). This line was found in Spatial Analysis in GIS Enviroment, and that is why I bring it up in the first place. Should we not bring up the rest of the world? Other parts of the world use GIS, and they could contribute to U. S. A.'s knowlege of GIS and vice versa. (Of course, we are already doing this but not to the extent we could be doing.) I know that many other countries have difficulty feeding their own people, so GIS appears to have little need in the country. I think if we showed them that GIS could be used to help feed their country, they would promote GIS. Another words when we consider spatial analysis, we should not only consider the technical sense of the word, but also the global implications of GIS. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1684) ----------------------------------------------------------------------- [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 07, 1998 10:20 AM Author: Guangxiang Cheng (gcheng1@tiger.lsu.edu) Subject: Answer four questions 1. yes 2. geostatistics 3. example with data analysis. For example, in spatial analysis, handling the same data with different methods of different spatial analysis can really help the virtual seminar participants to learn a lot. 4. if example presented, my interest can be more triggered (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1767) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 04, 1998 10:02 PM Author: Mihye Bark (gabark@unix1.sncc.lsu.edu) Subject: Report for Spatial Analysis in a GIS Environment Geog.7973-Advanced GIS Seminar (Report for Spatial Analysis in a GIS Environment) Spatial analysis is one of the most important functions of Geographic Information Systems (GIS). But the lack of integration of spatial statistical procedures and spatial models is still perceived as a major shortcoming. The spatial analysis in a GIS environment can be understood by identifying basic concepts, a variety of methods, techniques, approaches, problems, and future researches for the analysis of spatial and time-space data and models. Under UCGIS, the term "spatial analysis" includes a wide range of techniques for analyzing, computing, visualizing, simplifying, and theorizing about geographic data. Spatial analysis ranges from simple description to full-blown model-driven statistical inference. Originally, spatial analysis was an approach to the study of quantitative and statistical geography which emphasized the locational (point, lines, areas, and surface) patterns of a variable or series of variables (in the 1950s). It should not be assumed that the existing spatial analysis techniques derived from quantitative geography or spatial statistics of the 1960s is appropriate for application in GIS. The problem has been in identifying the technologies, procedures, and methods to provide basic analysis function that are truly useful and needed to GIS. For the appropriate spatial analysis in GIS, analytic and computational methods, new forms of statistical analysis and theories are needed. In GIS, data are divided into spatial data and non-spatial data (the associated attributes). But spatial analysis within GIS has been emphasized under recognition of spatial data. Enormous amounts of data of earth environmental, demographic, social and economic aspects are becoming available at different spatial and temporal scales. Yet our ability to test and develop theories, to extract meaning, and make useful decisions from the data has not kept pace. Traditionally, GIS is considered to perform four basic functions on spatial data: input, storage, analysis, and output (Goodchild 1987). Analysis function (Anselin and Getis 1992) includes (a) data selection (which considers the sampling of observational units from the data base and the choice of the scale of analysis), (b) data manipulation (which encompasses the partitioning, aggregation, overlay, and interpolation procedures needed to convert the selected information into meaningful maps and surfaces), and (c) techniques and methods of spatial analysis. Generally, the techniques of spatial (data) analysis are divided into two major categories (exploratory and confirmatory spatial data analysis), although many techniques incorporate aspects of both. Exploratory (data-driven, or inductive approach) spatial data analysis (ESDA) involves seeking good descriptions of data, thus helping the analyst to develop hypotheses and models about such data. On the other hand, confirmatory (model-driven or deductive approach) spatial data analysis (CSDA) is to test and confirm the hypotheses, develop methodologies and theoretical models, then go on to collect data. This would include most of the "traditional" techniques of spatial data analysis, such as hypothesis tests, estimation of spatial process models, simulation, and prediction. In ESDA one searches for structure and association, while in CSDA one evaluates the evidence. Most researches, however, are dependent on explanatory and theoretical aspect of GIS, not practical and confirmatory technology of GIS. Important applications in spatial analysis encompass disease distribution, traffic management and land use planning, environmental problems, landscape characterization and measurement, social, cultural, and economic analyses, physical process, improving the accessibility and equity of opportunities and services. For the development of spatial analysis in a GIS environment, We must to clarify several fundamental issues and problems, and thus develop new forms of tools, methods, and models. They are to (1) have a better understanding of spatial scale (or modifiable areal unit), boundary effects, spatial interpolation, spatial sampling procedures and incomplete data, spatial interaction, spatial association, spatial heterogeneity, and spatial movement, (2) develop method for handling massive spatial data at the different spatial and temporal scale (e.g. disaggregated census data, remotely-sensed data at a global scale). Also methods of spatial sampling procedures and data manipulation must be improved. In data manipulation, no single method of spatial interpolation may be single out as 'best'. Rather, the 'selection of an appropriate interpolation model depends largely on the type of data, the degree of accuracy desired, and the amount of computational effort afforded, (3) extend exploratory methods of analyzing spatial and space-time data and develop confirmatory (significance) procedures to test hypotheses, (4) develop appropriate models in a GIS environment (eg. econometric modeling and spatial interaction models under aggregate versus disaggregate, simple versus complex, and top-down versus bottom-up approaches), (5) develop operations research and computationally intensive procedures for the use of large data sets and the solution of complex spatial phenomena (e.g. neural nets, fuzzy sets, wavelets, microsimulation, artificial intelligence, natural language processing of textual information, artificial life, real-time data analysis, numeric optimization techniques and massively parallel algorithms), (6) develop the scale correctives or independent methods to analyze the impact of scale and study global versus local effects, and (7) discover the appropriate GIS tools for pattern recognition and analysis, data generalization, edge detection, and extend use of geostatistical procedures (variogram and kriging). Finally, the integration of spatial data, spatial analysis and modeling capabilities and evaluation models into computer-based Decision Support Systems (DSS) is desirable. Future success of GIS will illustrate the potential benefits of GIS technology as a decision supporting tool for spatial analysis. To do so, GIS researchers must develop appropriate (spatial) analysis function that are truly useful and needed to GIS environment and standards for their application. Also, those who are well-schooled in spatial statistics, geostatistics, spatial econometrics, mathematics, geocomputational algorithms, data collection, data manipulation, data structure and time-space modeling, programming languages, theories of computation, and computer technology will be in the best position to make advances in spatial analysis of GIS and influence the next generation of decisive research undertaking. Reference Anselin, L., 1996. SpaceStat, Version 1.80. Morgantown, WV: Regional Research Institute, West Virginia University. Anselin, L., and A. Getis, 1992. Spatial statistical analysis and geographic information systems. Annals of Regional Science 26:1933. Cressie, N., 1991. Statistics for Spatial Data. New York: Wiley. Fischer, M. F., and P. Nijkamp, editors, 1993. Geographic Information Systems, Spatial Modeling and Policy Evaluation. Berlin: Springer Verlag. Fishcer, M., H. J. Scholten, and D. Unwin, 1996. Spatial Analytical Perspectives on GIS. London: Taylor and Francis. Fotheringham, S., and P. Rogerson, editors, 1994. Spatial Analysis and GIS. London: Taylor and Francis. Fotheringham, S., and P. Rogerson. 1993. GIS and spatial analytical problems. International Journal of Geographic Information Systems 7(1):3-19. Longley, P. and M, Batty, 1996. Spatial Analysis: modelling in a GIS environment. Cambridge: Geoinformation International; New York: Wiley. Miller, H. J., 1996. GIS and geometric representation in facility location problems. International Journal of Geographic Information Systems 10(7):791-816. Openshaw, S., 1994. Two exploratory space-time attribute pattern analyzers relevant to GIS. In S. Fotheringham and P. Rogerson, editors, Spatial Analysis and GIS. London: Taylor and Francis. Openshaw, S., 1996. Parallel simulated annealing and genetic algorithms for reengineering zoning systems. Geographical Systems 3:201-220. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1672) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 07:36 AM Author: Jay Raiford (jraifor@lsu.edu) Subject: Spatial Analysis Report - LSU Attachments: Spatial Analysis.doc (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1674) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 08:18 AM Author: Kurt L. Johnson (Kurtljohnson@worldnet.att.net) Spatial Analysis in GIS 2nd try Attachments: SPATIAL1.doc Spatial Analysis in GIS by Kurt Johnson Spatial analysis is a recently developed tool used in geography. Its purpose is to help us understand more about the interactions of our surroundings. Generally speaking, our surroundings consist of many different kinds of objects referred to as features. These features can can be graphically depicted as points, lines, or areas whereas, the information describing these features are referred to as spatial data (DeMers, 1997). The data is feature specific and generally gives some insight about the feature's location, attributes, and relationship to its surroundings. To make sense of spatial data, the data requires analysis to determine just how a feature is influenced by its surroundings or how the surroundings influence the feature. The analysis should be a compilation of the representatively organized data, facilitating the extraction of useful knowledge. Initially, this might seem simple but as more and more data sets are considered, the task of understanding how everything is related becomes very complex. Therefore, spatial analysis must proceed within an organized conceptual framework allowing us to understand the interrelatedness between individual features and events (Bonham-Carter, 1994). The framework described here is referred to as a Geographical Information System (GIS). Before any form of spatial analysis can be applied, some attention needs to be given to spatial data quality issues. Because of recent technological advancements spatial data is being collected at ever increasing rates and levels of accuracy. This conforms well with society=s increasing demand for more exact and detailed data. However, the down side of this is that the increased demand for spatial answers opens the door for questionable levels of quality control and quality assurance. Prior to computer advancements, cartographers were principally responsible for using spatial data to produce a product, namely a map. These processes included data collection, data editing, map compilation, and map reproduction (Guptill and Morrison, 1995). In this sense, cartographers could be considered both data producers and data users, but the end user would be the map users. With today=s powerful computers and advanced software programs, the processes traditionally performed by cartographers are now being utilized by other disciplines. Essentially, technological advancements have not only increased the number of spatial data producers, but more importantly it has greatly expanded the potential for spatial data users to employ cartographic processes (Guptill and Morrison, 1995). Although these technological advancements are progress, they come with a price and that price is a reduction in control. The end result of our advancements surface in the concerns about product quality issues. These issues arise from cartographic processes and the increase in numbers of players utilizing those processes. For those individuals not well trained in cartography, each decision about whether to use one data set over another can compound errors. Especially when considering many related features with diverse data sets. Much of the data that is available has been collected for a specific purpose and therefore the potential exists that many spatial data sets reflect a bias towards that purpose. This is an inherent problem of any data set, whether its intended use is for spatial analysis or not. Another problem stemming from today=s sophisticated computers are its rendering capabilities. Computers are very precise and methodical tools capable of transforming data into a visual products that look accurate, regardless of how accurate the data was initially (Guptill and Morrison 1995). This is a difficult problem to discern because a well rendered computer visualization has very few clues indicating the quality of the data. This problem might be a result of an intentional misrepresentation or one resulting from a computer operator=s skill being greater than his knowledge of the data. With these concerns mentioned, its time to place spatial data in the context of GIS and analysis. A GIS is defined as a computer system for managing spatial data. It has functional capabilities for data capture, input, manipulation, transformation, visualization, combination, query, analysis, modeling and output (Bonham-Carter, 1994). These capabilities are achieved through computer software programs providing access to these various functions. At this point it might be helpful to dissect GIS word by word; The word geographic implies a location (in latitude, longitude coordinates) is either known or can be calculated. The word information implies that the data in a GIS is organized and structured to yield useful knowledge. The word system implies that GIS is comprised of several interrelated and linked components with different functions (Bonham-Carter, 1994). These three elements form GIS=s analytical framework. Its this framework that provides us the ability to look at our surroundings analytically. Therefore, the ultimate goal of GIS is to bring all three elements together facilitating an intelligent decision making process (Bonham-Carter, 1994). However, intelligent decisions are highly dependent on correct and accurate analysis of the spatial data. Analysis of spatial data is the essence of GIS. According to Pergamon 1994, analysis is the process of inferring meaning from data. Aronoff 1989, states that spatial analysis is what sets GIS apart from other information systems. The many different analytical functions described are too numerous to mention here but generally they consist of measurements, statistical computations, and modeling. Three themes suggested by Gatrell 1983, for utilizing these analytical functions are spatial arrangement, space-time processes, and spatial forecasting. The first theme, spatial arrangement refers to the locational pattern of the features studied. Space-time processes are concerned with how spatial arrangements are modified by movement or spatial interaction. Spatial forecasting incorporates time into the prediction of future spatial arrangements. Gatrell 1983, cites that these three themes are commonly found in introductory textbooks and that they are all linked to the study of spatial systems. The brief discussion above describes both conceptual components and concerns about spatial analysis in a GIS. Although the analytical concept of GIS was discussed, very little was mentioned about the methods, techniques, and approaches used to perform such analysis. This is partly due to my inexperience of using such analytical tools and to the problems that I have already encountered while using spatial data. Many of these problems have preempted any form of analysis. The first problem is in determining what scale of data will facilitate a useful analysis and if that data even exists or will it have to be collected. The next problem is how to use the many different formats of data so that they can utilized. This involves concerns about transformation; what is lost in transformation or what is transformed in error. A third problem is the complexity of GIS programs, their capability of handling the data sets, and if they will perform the analytical functions desired. I can appreciate the research goals for this topic and find them intriguing but its hard for me to conceive of these goals when so many of the problems I mentioned above hinder achieving spatial analysis. The white paper states that our ability to extract meaning from and make useful decisions has not kept pace. This is true however, I feel that if we don't establish a standardized and fundamental approach to using data, there will always be a significant amount of hidden error within the analysis. The users of GIS need to have a good understanding of how various data sets and their transformation into compatible formats will affect the analytical results. A comment from Berry 1995, states that much of our scientific knowledge lacks the spatial specificity in variable relationships. Geographic information systems are the tools used to characterize these relationships but we need to improve our understanding of how we can express these relationships more representatively. Aronoff, S., 1989. Geographic Information Systems: A management Perspective. WDL Publications, Ottawa, Canada. DeMers, M.N., 1997. Fundamentals of Geographic Information Systems. John Wiley and Sons, Inc. New York, NY. Berry, J.K., 1995. Spatial Reasoning for Effective GIS. GIS World Books, Fort Collins, Co. Bonham-Carter, G.F., 1994. Geographic Information Systems for Geoscientists: Modeling with GIS. Pergamon ( Elsevier Science Ltd.), Oxford, U.K. Gatrell, A., 1983. Distance and Space: A Geographical Perspective. Oxford University Press, New York, NY. Guptill, S.C. and J.L. Morrison, ed. 1995. Elements of Spatial Data Quality. Elsevier Science Ltd., Oxford, U.K. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1675) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 10:23 AM Author: Doug Albert (dalbert.unix1.sncc.lsu.edu) Subject: Summary of digital earth and five research priorities Doug Albert Geography 7973 Report 1 1. INTEROPERABILITY OF GEOGRAPHIC INFORMATION "The term interoperability refers to a bottom-up integration of existing systems and applications that were not designed to be integrated when they were built." (UCGIS, 1996: 122) It is difficult to transfer, access, or control data from one system with another system. Interoperability can help resolve these problems if one system is able to understand another system. This is called semantic interoperabilty. Work is being done on storing and representing metadata. In the future selecting and updating specific metadata will become increasingly important. Interoperability will become indispensable in environmental modeling. 2. THE FUTURE OF THE SPATIAL INFORMATION INFRASTRUCTURE "Despite large investments in geographic data development by government and private sectors, there is often a lack of knowledge and experience of the complex policy - related issues that arise from the community-wide creation, compilation, exchange, and archiving of large geography data sets." (UCGIS, 1996: 124) The government sector develops much of the spatial information infrastructure, while the public makes certain of governmental accountability and their democratic decisions. UCGIS proposes four large areas in which research can help in the future of spatial information infrastructures: Information Policy, Access to Government Geographic Information, Economics of Information, and Local Generation and Integration of Geographic Information. 3. UNCERTAINTY IN GEOGRAPHIC DATA AND GIS-BASED ACTIVITIES Geographic data has three attributes: typological, locational and spatial. Uncertainty is found in these attributes. The attributes can only approximate reality, and these same attributes can change over time. This creates discrepancies, "and may be further amplified by spatial data management and analyses in a GIS environment". (UCGIS 1996, 126) This is called uncertainty. Many times GIS data is assumed to have no uncertainties. Uncertainty is found at every phase of the geographic data's life and the user's can create uncertainty. More research into GIS uncertainty is needed. 4. SPATIAL DATA ACQUISITION AND INTEGRATION Accuracy and resolution is increasing in geographic data, but this data must be integrated with other data. A digital orthophoto quadrangle (DOQ) is an example of this integration. Accuracy is also very important. There are two types: relative position and absolute position. The relative position is usually known, but the absolute position is poor. Edgematching is also a problem. Consortium agreements and communication technologies such as the internet have also increased data integration problems. Tools need to be developed for data integration based on good quality research. These tools are referred to as conflation. Conflation requires specialization and collaboration of many sciences. Future developments will include field GIS, a GIS system that can be taken directly into the field. 5. COGNITION OF GEOGRAPHIC INFORMATION GIS can only realize its full potential by recognizing certain parts of the human cognition. GIS must mirror a person's natural thought process. Hopefully GIS will become more assessable to inexperienced and disadvantaged users, and become even more effective for experienced users. Education will be improved. In-vehicle navigation systems (INVS) has used and will continue to use human cognition to improve its effectiveness. The future of map libraries on the internet "depends on our own ability to provide a user interface that successfully reproduces all of the map library's functions". (UCGIS 1996: 121) Many subjects will need to work together to promote cognitive aspects in GIS. 6. THE DIGITAL EARTH The world is collecting lots of information on the earth's surface, but much of it goes unused. Why? The information is displayed in a way that is not easily recognized by humans. Even the Macintosh and Windows operating systems have this problem. A 'Digital Earth'is needed. "A multi-resolution, three-dimensional representation of the planet, into which we can embed vast quantities of geo-referenced data." (Gore 1998: 1) For example a small girl could take a virtual field-trip to Paris. It would take hundreds of thousands of people to help create the digital earth for both public and private access. Some of the technologies that are needed include: computational science, mass storage, satellite imagery, broadband networks, interoperability, and metadata. Present applications include conducting virtual diplomacy and fighting crime. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1680) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 10:30 AM Author: Doug Albert (dalbert.unix1.sncc.lsu.edu) Subject: Spatial Analysis in a GIS Environment Doug Albert Geography 7973 Report 2 SPATIAL ANALYSIS IN A GIS ENVIRONMENT "The term 'spatial analysis' encompasses a wide range of techniques for analyzing, computing, visualizing, simplifying, and theorizing about geographical data." (UCGIS 98: 15) These methods can be simple or complex. Remote sensing is providing more data on the earth, but finding meaning and making informed decisions from the data have lagged behind the copious amounts of data. New statistical analyses with new theories have to be developed in order to find the relationships between variables at different resolutions. Spatial data is different from other types of data. One must find relationships between the variables under study and even the spatial units themselves. When studying spatial data, one needs to ask questions on spatial association, spatial heterogeneity, spatial scale, unfinished data, and boundaries. If these answers are not answered, GIS will never be fully developed. On the other hand, there have been new developments in spatial analysis. For example the cost of analyzing spatial data has come down, and many conversions between analyzing and modeling at different scales are being produced. Data models and GIS are linked by spatial analysis resulting in applications being improved and research findings having more depth. GIS includes georeferenced data and the tools that handle the data. Spatial analysis is helping applied fields. UCGIS believes that spatial analysts in the human and physical sciences need to take advantage of GIS. Data specialists and computer scientists can make advances in the field. Combine these people with applied scientists, and one is in a better situation to take on future research projects. The United States needs to advance analytical techniques in order to stay on top of GIS technology. Other countries or the World itself need to be included in the advancements of GIS. Can not the countries of the world advance as one in GIS technology. The handling of large data sets must be improved upon, so than one can make better use of GIS tools. The United States does have access to large amounts of data, but much of it goes on used. There are many benefits in improved spatial analysis and GIS. Disease distributions can obtain a spatial and temporal framework. Real-time traffic analysis can improve traffic problems. GIS can help analysis the data obtained from environmental models. Classification of land cover and land use can be improved upon by GIS. Social scientists can study social problems with data that has been analyzed. GIS can be used with hydrologic and climatic data to determine global change. Hopefully GIS and spatial analysis can help disadvantaged population gain access to opportunities and services. The future of GIS must not only be spatial but also spatial-temporal. The handling of large amounts of data must be improved upon so that data can be used to its fullest. Models that represent reality need not only a spatial framework but a space-time framework. Variogram and kringing must be used and developed more in GIS. The affects of scale need to be analyzed more. Global studies lack fine resolution. Local measurements need to be used in order to make global studies more useful. Procedures need to be created that identify extreme or important events. GIS must develop with the advances in computational abilities. Econometrics modeling must be used more in GIS. Spatial interaction models should be developed in GIS. Operations research ought to be used to a greater detail. (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1681) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 10:47 AM Author: Guangxiang Cheng (gcheng1@tiger.lsu.edu) Subject: Spatial Analysis Report (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1682) ------------------------------------------------------------------------ [Top][Previous][Next][Print][Reply][Edit*][Move*][Delete*] Date: October 05, 1998 10:50 AM Author: Guangxiang Cheng (gcheng1@tiger.lsu.edu) Attachments: Geography7973-Report2.doc (http://forums.library.orst.edu/forums/Index.cfm?CFApp=7&Message_ID=1683)