February 17

• The usefulness of GIS increases with the growth of other geospatial technologies because GIS integrates and organizes data from all of them. However, geographic data is often inconsistent, redundant, and/or ambiguous, because different users represent the same data differently, depending on their point of view and needs. For example, a satellite photograph may show the accurate shape of an object, a tourist map may show a less accurate shape for that same object but include its name, and a database may list the object's name and that of its architect.
       That's where we turn to conflation, the challenge of combining information from disparate sources in such a way as to retain accurate data, eliminate redundancies, and reconcile data conflicts. If we handle them correctly, different representations of the same geographic reality give us richer information.¹
       That, as I understand it, is the theory. So, how does it really work in practice? What techniques and software packages are best for the job? Which GIS shop is doing the most sophisticated and effective conflation work? What are some the most impressive case studies?

• As an increasing portion of GIS input is remotely sensed data, GIS users are paying more attention to the uncertainty in remotely sensed image classification. This uncertainty can arise from mixed pixels, poor class definitions, and transition zones (fuzzy boundaries).²
       To what extent does the uncertainty reduce the value of remote sensing as a GIS data source and affect the return on investment (ROI)? What are the most promising approaches and tools available to minimize the uncertainty?

• GRASS, the Geographical Resources Analysis Support System, was popular in the 1980s and the early 1990s, when there were comparatively few software packages that could justifiably call themselves Geographic Information Systems. GRASS was a UNIX-based raster GIS used widely in both government and academe. First developed in 1982 by the U.S. Army Corps of Engineers' Construction Engineering Research Laboratory (CERL), GRASS was then the only public-domain GIS, well before the use of Open Source software became widespread. CERL made the source code for the entire system public and encouraged the "GRASS community" to develop and add its own code to the package. This led to the formation of the Open GRASS Foundation, which then became the Open Geospatial Consortium, Inc. (OGC).
       GRASS was able to incorporate new developments in geographic information science, such as surface analysis and map algebra, without having to wait for commercial software developers to incorporate them into proprietary products. In 1995, however, CERL discontinued its support of GRASS and its use declined sharply. Recently, as the Open Source model of software development has experienced a renaissance, so has GRASS.³
       What are the prospects for GRASS and other Open Source GIS software development? Who's doing what? Is Open Source GIS code more robust and flexible than commercial products? Are commercial products better for certain applications or environments?

• Many countries are developing Spatial Data Infrastructure (SDI) to manage and use their spatial data assets more efficiently, to facilitate and coordinate the exchange and sharing of spatial data between stakeholders in the spatial data community, and to assist in decision-making that has an important impact within their national boundaries. Over the past few years, many countries have begun to build National SDI (NSDI), starting with its main element, a national spatial data clearinghouse to serve as the access network of an NSDI that facilitates access to the spatial data.⁴
       How are national clearinghouses developing? How practical are they? Who uses them and for what purpose? Do they include efficient Web interfaces? Are they adequately funded? Obviously the answers to these questions will vary widely from country to country.

• Millions of geological, biological, and cultural specimens housed in natural history museums are cataloged with very imprecise descriptions of where they were found.⁵
       What is the best way to georeference the source of these specimens, so that scientists can study their geographic patterns? Who's doing it? Will they ever catch up with the enormous backlog?

• The Internet and, in particular, the World Wide Web, have broadened the use of powerful GIS tools beyond a small band of users and now GIS offers a more populist environment, with a wider range of applications and services. This symbiotic relationship between the Internet and geographic information services allows much greater dissemination of geographic data and analysis. The challenge, of course, is to encode, transport, and promote data interoperability over heterogeneous Internet environments. To this end, the development of Geography Markup Language (GML) and the work of the Open GIS Consortium and the International Standards Organization (ISO) are crucial. ⁶
       What are the remaining obstacles to fully Web-enabled GIS?

• With the advancement and convergence of GPS, the Internet, and wireless communication technologies, mobile GIS is rapidly becoming the preferred method of data acquisition and validation. Mobile GIS brings analysis from the laboratory right to the point of data collection, thereby transforming the way data are collected in the field. It allows field scientists to increasingly focus on their research, rather than on the technical details of data collection, and to gather complex digital data, such as GPS data and digital photography, while also accessing pre-existing digital databases. Digital instruments that communicate through local wireless systems, such as Bluetooth, help integrate data from various instruments into a single spatial database through mobile GIS.
      &nbspThe remaining obstacles to a full implementation of mobile GIS include the short communication range of wireless networks; the requirement for broad bandwidth; the limitations of the map display and user interface design of mobile GIS applications; and the limited integration into mobile GIS of Internet mapping technologies originally designed for desk-top clients and standardized Web browsers.
       Meanwhile, how are the pioneers of mobile GIS doing? For what applications is the current technology sufficient? Who could benefit most from expanded capabilities for mobile GIS? Who will?

• Am I asking the right questions?

Department of Corrections
.