Risto Päivinen is Director of the European
Forest Institute, Joensuu, Finland
and Coordinator of the International
Union of Forestry Research Organizations (IUFRO)
Task Force on Global Forest Information Service.
Roger Mills is Head of the Library and
Information Service, Oxford Forestry Institute,
Oxford, United Kingdom.
Michael Hailu is Head of the Information
Services Group, Center for International Forestry
Research, Jakarta, Indonesia.
Jarmo Saarikko is Information Manager
at the Finnish Forest Research Institute,
Helsinki, Finland.
The continuing revolution in information and communication technology raises questions of reliability, transparency, privacy and accessibility of information.
Trees grow more slowly than crops - during the last century this simple fact dictated a path of largely separate development for forest-related information retrieval within the broader field of agricultural, environmental and biological information. Specialist services have emerged for those seeking data on trees and forests, in recognition of the long "half-life" of literature on this subject and the need to preserve experimental data for many years in order to make valid comparisons and to avoid unnecessary repetition.
At the beginning of the twenty-first century, technology offers unparalleled opportunities to access and simultaneously interrogate such information resources, which have been collected from all over the world. The mass of information has become both larger and less mysterious in a very short time; but now, more than ever, we need signposts through the myriad paths that are opening up and guidebooks to the many destinations they lead to. Fortunately, the library and information world has long experience of such signing and guiding and there is much to build on.
Arranging books in a logical sequence dates back to antiquity, although it was the invention of printing that made the need to do so a "mass-market" issue. Printing spawned a rash of libraries across Europe which ensured that information stayed where it was wanted by literally chaining the books to the shelves. The growth of large national libraries and the public library movement of the nineteenth century led to the development of coded universal classification schemes such as the Dewey Decimal Classification, and over the following 100 years these became the principal information retrieval mechanism in many subject areas, including forestry.
Foresters seem to be orderly by nature, and the International Union of Forestry Research Organizations (IUFRO) created a bibliographical committee early in the twentieth century which addressed many of the key issues of the time and led, either directly or through the enthusiasm of its members, to the establishment of Forestry Abstracts (now published by CAB International), the Oxford Decimal Classification for Forestry (ODC) (now the Forest Decimal Classification, FDC) and the concept of "libraries of deposit" for forestry literature in the United Kingdom and Russia. The advent of computers in the 1970s revolutionized the production of indexing and abstracting services, and the book trade became one of the first to adopt the now ubiquitous "article numbers", which are now bar-coded on to items of every description, when it adopted the International Standard Book Number (ISBN) and International Standard Serial Number (ISSN) as unique identifiers. The quantity of material published continued to grow annually (and is still growing), but microfilm, photocopying, fax, video, CD-ROM and e-mail greatly simplified archiving and distribution. For a while it looked as if the battle for "bibliographic control", as it had come to be known, was won.
Then Tim Berners-Lee invented the World Wide Web and immediately the battle reopened on a new front. For the first time it became simple for authors to bypass conventional publishing routes and communicate directly with their intended audience. Existing bibliographic control mechanisms were, however, largely geared to the products of conventional publishers. Suddenly new rivers were flowing across the information floodplain, although it was still difficult to see what these rivers contained until the arrival of the commercial search engines. These were like huge nets that dipped into the new rivers so we could see what they contained at that moment, but not where it had come from, where it was going and what it might consist of. These three questions - provenance, sustainability and quality - have become the key issues for twenty-first century information retrieval and the focus for forest information mechanisms today.
Even the most effective search engines are able to find only one out of every 500 Internet sites (compared 6 June 2000)
- Source: SearchEngineWatch.cpm
Since anyone can now publish on the Web with little if any editorial control, the "buyer" has to be more wary than ever and in a way that has not yet become familiar. Users of information have always needed to evaluate its quality - including its relevance for their own purposes, the appropriateness and accuracy of the methods and measurements used, its timeliness, the compatibility of definitions, and the degree to which these aspects are revealed in the information itself - but the Web has made such evaluation more complex.
What is the author's background? Is his or her work reliable? What do others in the field think of it? Is it politically motivated, and if so is that motivation clear? Carrying out the research to answer these questions can be time-consuming for the end user. In the world of print, readers could rely on publishers and reviewers to do it.
Will the information still be available when it is wanted again: tomorrow, next year, next decade? If it resides only on an individual's personal computer, the answer is almost certain to be no, and much effort is now being expended in many countries to set up major national archives for digital information to ensure its long-term preservation and accessibility. However, only a fraction of what exists on the Web can be permanently archived, and selection requires evaluation and quality assessment, a process that can never be wholly automated. Collection development policies, which have always been necessary for printed materials, are now needed for digital information.
Recent evaluations (www.brightplanet. com) suggest that, altogether, perhaps 550 000 million documents are accessible on the Internet, but general search engines index only about 1 000 million of them (see Figure). The unseen "deep Web" is thought to include around 100 000 content-rich searchable databases - 95 percent of which are publicly accessible without a fee. New software can make these databases "visible". Specialist subject-based filters using generic search engines can limit retrieved sites to manageable numbers based on relevance and quality assessment. However, some search services give prominence to sites that have paid for their listing style; this is inevitable as publishers and information providers seek new methods of generating income to replace the print packages on which they have depended.
Is there a need to package information by subject as journals have done? The electronic journal as a reproduction of a print version may have a limited life. The development of managed electronic "communities", where creators, distributors and users of information interact in a common forum, may offer clues as to the shape of electronic publishing in the future.
The issue of how to index the largest library ever known is currently being debated. Should it be done by machine, using artificial intelligence, knowledge representation or agents, or by distributed metadata, using subject gateways? A true catalogue creates relationships among documents that may not appear to be related, by categorizing, identifying, connecting, choosing and discarding. At present human brainpower is needed for this - but will this always be so?
A global positioning system (GPS) device
- FAO/20510/A. PROTO
Predicting technological development is a difficult task. Few foresaw the changes that personal computers and the Internet were to bring about with such rapidity over the last 20 years. Networked computing will change the way we work even further; many more technologies will appear to exploit the increases in computing speed and storage capacity, combined with ever-decreasing equipment size.
In the near future computing will become more personal and mobile, and thus the number of points where data can be entered will increase dramatically. Estimates suggest that the number of Global System for Mobile Communications (GSM) telephone subscribers will double from 300 million to 600 million within the next three years. Operators are already providing new services such as Wireless Application Protocol (WAP) and General Packet Radio Service (GPRS), although the handsets using these techniques are only just beginning to appear on the market.
GPRS phones provide continuous on-line access to e-mail and the Internet, and their access speed can be up to ten times faster than that of regular data transmission over GSM. The benefits are connectivity and immediacy: information can be sent and received at any time without the need for dialling or logging on. Another advantage is localization: subscribers can receive information related to their location, such as weather reports, navigation help, information on nearby services or general data. The current disadvantage of the small screen of the telephone handset display will be overcome by new hand-held communications terminals combining a phone, a Global Positioning System (GPS) receiver and a small personal computer.
>Both WAP and GPRS are still stepping-stones for telecommunications operators who are already planning new third-generation mobile networks, such as the Universal Mobile Telecommunications System (UMTS). These networks will use megabyte-level high-speed transmissions, enabling video-quality reception on hand-held equipment or full on-line usage of Geographic Information System (GIS) applications (UMTS Forum, 2000). They may be in operation in a few years' time. Market penetration of such new services will increase as prices fall. Applications in forestry may include mobile office functions and corporate Intranet access.
Many forestry workers operating alone already carry portable phones as a safety measure. In Finland, for example, one-third of all emergency calls come from mobile phones. Often the mobile callers cannot give an exact location. New locating technologies will be used on hand-held phones to make automatic location detection possible. GPS are now routinely used in some forest operations; they have improved the logistics of wood transport, for example. A GIS system installed on the harvester machine is used to establish the location of wood piles. The driver of the harvester transmits the location to the transport company, which sends the information to a truck with a GPS receiver combined with a GIS system on a personal computer. The driver is then able to retrieve the pile from a distant, previously unknown location deep in the forest.
For military reasons, the United States used to distort GPS satellite transmissions, but the distortion signal was removed in May 2000. This has made location much more accurate, from within 100 m to within 10 m, and made the national correction signals provided in some countries unnecessary. As the United States still reserves the right to distort GPS signals, the European Union (EU) is developing its own service, Galileo, which is mainly directed towards civilian use.
For a forester of the future with a mobile hand-held terminal, discovering plot information in the field will become much easier. There will no longer be a need to carry all the background information, and further information can be appended to the networked data warehouse on the move. As more data become available on the network it may be possible to check land ownership, management plans and certification status and to retrieve geological and biological information on the spot by sending the location, which will be pinpointed automatically with a satellite system. The inquirer may be able to access systems where many diverse clients are able to store observations (such as plant and animal observations in regard to biodiversity). If the queries are prepared in advance for a particular location, the retrieved data can be entered in a portable GIS system or printed as paper maps to take into the forest.
As an example, a forest owner wishing to obtain bids for wood lots could do so from the forest by using a portable GPRS phone to connect to an on-line forest products exchange through which he or she could post a sale announcement or check who is buying that kind of wood. Buyers would have automatic agents to check new announcements, and if the type, quality, location and price match what a buyer requires, the agent could automatically send an offer to the owner. After choosing from among the offers received in this way, the owner would use an electronic identification and signature to make the deal on-line. To contract a harvester, the owner would then transmit the coordinates of the wood's location, established by GPS, directly to the harvester's GIS system.
Localization services on mobile networks reveal the user's location to others. There is always a danger of "Big Brother" surveillance, which may cause some users to avoid this kind of service. However, software engineers are developing encrypting and anonymizing software to avoid such dangers.
The stream of incoming data is increasing as high-resolution satellites become much cheaper and available to civilian use. Many space launching facilities, which used to be strictly military or governmental, are now also launching commercial cargo. New images of spy-class accuracy with resolutions as precise as 1 m2 are already available for general users at affordable prices.
With improving satellite technology comes an increasing demand for satellite-based earth observation networks to monitor the environment and development. In the forest sector, satellite observations will be required for monitoring of global vegetation, deforestation, afforestation and reforestation. Point observations on the ground and their analysis are labour-intensive and expensive, and the point network is often too sparse. Automatic observation stations will be developed for various types of environmental observations, including images, which will then be transmitted over mobile networks. The combination of automatic ground observations with very accurate satellite data will result in far more accurate predictions and models, which will then lead to better decision support and expert systems. These will be developed into more user-friendly graphical map-based GIS tools.
Adapting to these new technologies will not be without problems. As computing becomes more ubiquitous and easier to use, collaboration, asking questions, exchanging experiences and sharing information, regardless of location, will all become cheaper and more efficient. However, mobile networks are subject to the same social and behavioural laws as any other human network. Non-facial and text-based communication is not easy in a multicultural group. New common rules will be needed in electronic communities, in order to handle situations where people do not behave as expected.
New technologies develop much faster than the applications that use them. While the network provides enormous opportunities for storing data, questions of data ownership and sharing are still a major obstacle to the wide usage of open GIS services. As mobile computing begins to provide more opportunities for storing data that are bound to a location, questions of the source and quality of data will once again be highlighted. How reliable is the source? How were the data measured? New electronic communities are developing social tools for evaluating opinions and articles, giving some sources higher credibility ratings than others. In the future, this approach, which mimics the word-of-mouth networks of colleagues and friends, may be adapted for the evaluation of scientific and other data.
In exploring how forest-related information and knowledge will be exchanged in the future, it is not possible to ignore the widening gap between developed and developing countries in their endowment of information and communication technologies.
Global Forest Information Service - improved access to information In order to improve access to quality information, the International Union of Forestry Research Organizations (IUFRO) has developed the concept of a Global Forest Information Service (GFIS) (IUFRO, 1999). GFIS is based on a distributed network of metadatabases which catalogue the information resources - digitalized or not - of contributing GFIS partners. (Metadata are data about data.) The Internet and the Web are used to facilitate the location and searching of metadata catalogues. GFIS will provide a standardized core of metadata (catalogue) fields, a standardized set of key words on which to search, and a standardized interface between Web sites and the databases. Participants should be able to use their existing metadata catalogues with little modification or with the aid of conversion tables. Furthermore, participants who do not yet have catalogues can immediately become compatible by following GFIS guidelines. The envisioned structure and information flow within GFIS is illustrated in the Figure. GFIS aims to be a source of reliable data. To ensure this, data providers will be asked to provide details of data provenance, updating arrangements, contact addresses, etc., which will allow end users to make comparative assessments of resources and decide for themselves the appropriateness of a given resource for their needs. GFIS will not censor data, but just as a reputable journal will not usually publish an anonymous article, it will not normally include resources if the relevant quality assessment metadata are not submitted. However, in areas where few resources are available, poor data may be better than no data, so a degree of flexibility will be built into the acceptance system, which will include regular reviews to ensure that standards continue to be met. For forest resource information, for instance, metadata describe the purpose and date of the inventory, the number of field samples used and the field quality control methods. From that information the user should be able to conclude whether the statistics of forest area and volume of growing stock are acceptably reliable. The forestry sector is not alone in the process of improving the link between information users and providers. The first Consultation on Agricultural Information Management, organized by FAO in June 2000, discussed ways of improving the capacities of decision-makers, professionals and the public at large to access and use the information that is essential for achieving sustainable agricultural development and food security. The consultation covered all aspects of agricultural information management, capacity building and improving access to relevant information (FAO, 2000). Structure of the Global Forest Information Service
1. GFIS information server: coordinates the housekeeping functions of the GFIS node network
(e.g. user registry, general information, discussion forums, list servers, gateway to distributed metadatabases). |
Globally, the information technology sector is expanding at twice the rate of the world economy. However, this trend has been largely concentrated in developed countries. For instance, compared with a North American and European average of one Internet user in every three people, developing regions fare poorly with only one in every 750 for Africa, one in 125 for Latin America and the Caribbean, one in 200 for Southeast Asia and the Pacific, one in 250 for East Asia, one in 500 for the Arab States and one in 2 500 for South Asia (Jensen, 2000; UNDP, 1999). Developing regions also lag way behind the United States and Europe in terms of the numbers of Internet hosts, personal computers, mobile phones and phone lines that are available (GIIC, 2000). Within the forest sector, differences are even more pronounced, since developing country foresters are often in remote locations which are last to be served, while many developed countries with a large head start have achieved near 100 percent coverage, as in Finland, for example.
If most publishers abandon the print media and increasingly concentrate on real-time electronic publishing through the Internet, as many expect will happen, developing country foresters could become even more isolated from the global knowledge marketplace than they are today. Although the number of Internet users has been increasing steadily in developing countries, only a limited number of institutions in those countries are using the Web to deliver significant quantities of information because of poor access for the general population (which limits the importance of Web publishing in these countries), limited Web publishing and browsing skills and the relatively high cost of Web hosting services. In addition, Internet access in developing countries is mainly confined to capital cities, which poses a problem for foresters, who work largely in rural areas.
Mobilizing scientific information to promote sustainable management of forests in Africa With the approval of a three-year project supported by the European Commission DG VIII, the establishment of five Global Forest Information Service nodes within Africa (GFIS-Africa) has begun. These nodes will facilitate access to, and dissemination of, scientific and technical information on forests. They will be located (subject to confirmation) in Ghana, Senegal, Kenya, Zimbabwe and Gabon (Szaro, Martin and Landis, 1999). One of their objectives is to help build the capacity to develop and manage Internet-based systems that facilitate broad access to research information on forests in selected regional research institutions in Africa. Another objective is to enhance integration and comparability of national data on forests throughout African countries in the African, Caribbean and Pacific Group of States (ACP). Two project staff members at FAO in Rome will have responsibility for developing the interfaces for GFIS-Africa and for helping to set up equipment on-site. Each node will be provided with equipment, training, the salary of an information specialist and operational costs. As the project is implemented, many new partners will be included to help develop the network. |
Another issue is that of language. English dominates the Internet; some 80 percent of Web site content is currently in English (d'Orville, 2000), a statistic closely reflected in forestry sites.
With concerted efforts by government, the private sector and civil society, the digital divide can be narrowed and the power of the Internet and the Web can be harnessed to bring information and knowledge to millions of users in developing countries. The United Nations (UN) has identified the following seven key areas where measures need to be taken to build a truly global knowledge society (UNDP, 2000):
Information is sought in many contexts. It may have entertainment value that is not directly utilitarian, but perhaps it is most frequently needed for decision-making. Seldom is all the desired information available; so decisions are almost inevitably imperfect. However, new information should enable improved decisions to be made.
The following are two examples of how new technologies and decision support systems may bring about change in decision-making over the coming decades. The first example illustrates the professional's point of view, the second the consumer's.
It is 2050 and, in general, the requirements for management planning remain the same as they were at the beginning of the century: a forest area needs to be managed in order to meet the goals of the decision-maker. However, now more information is available to a larger group of stakeholders, so there are more groups and individual citizens ready to comment on the management plan. The public also has better access to relevant information sources and is sometimes better informed on certain details related to the plan than are the professionals conducting the work.
A larger group of interested citizens are involved, meaning that the plan will have multiple, and often conflicting, goals. Advanced formulation of utility functions for different situations is therefore of great importance in reaching a balanced decision.
Forest inventory data are based on a combination of field samples, forest data collected previously and high-resolution remote sensing data. The amount of detail resulting from modelling these sources is far higher than could be produced in 2000, but the question of reliability is still relevant. Human interpretation of ecosystem elements in the field is still necessary, and the high cost of that part of the work impedes collection of the best possible data. This is especially true in the tropics, where field observations remain impractical in many cases. The fieldwork itself, however, has been facilitated by automatic positioning systems and on-line optimization of the measurements needed for accuracy.
The simulation of alternative scenarios is conducted in great detail and the forest is described almost tree by tree. In principle, growth models can utilize detailed information on soil, microclimate and tree competition, but in practice these data are often not fully available. Thus uncertainties remain, but better methods are available for evaluating their consequences in the decision-making process.
Evaluation of the alternatives is comprehensive and complex. Decision-support tools based on modelling and artificial intelligence are used to shortlist acceptable decisions. When selecting the best action, conventional optimization methods are still used. Advanced visualization (see Box on Using computer technology to predict the future, page 22) and virtual reality techniques enable more structured participation of non-professionals in the decision-making process. The chain from computers and consultants to decision-makers and other stakeholders has been enhanced by new software applications, a necessity since in many cases - including planning the management of public forest - interest groups are well informed about the issues involved and about similar cases in other regions.
Whether the management plan is better justified and accepted than it would have been in 2000 is questionable. It is more detailed and advanced, but at the same time the stakeholders are more aware of the possibilities, and their requirements may have grown faster than the ability of professionals to justify their actions.
This is a far simpler situation. There is only one decision-maker (the consumer) and the only variables are the quality of the table and the price, together with the environmental impact of production, which is of value to the consumer.
How are the environmental values associated with the production process to be evaluated in deciding between two tables of the same price and same quality? A handy WAP phone helps the consumer to locate the producers' Web sites and environmental statements. They look convincing. Then the consumer searches the Web for any complaints purchasers have made about the products and screens the blacklists of environmental groups - no problems in those sources either. The final decision is therefore based on the design.
In coming decades, consumers are likely to become more interested in the production process, as new technology makes it easier to access comments other than the producer's.
Predicting the future is risky, especially in the field of information technology, but the following conclusions seem reasonable:
Using computer technology to predict the future A sequence of close views generated by computer graphics showing the temporal development of a pine regeneration area (on the right in each image): from left to right, 1998, 2003 and 2008. Animations describing the details on the dynamics of the trees and whole ecosystem are under development.
Computer graphics may simulate spatial, temporal and seasonal variation as well as atmospheric effects: from left to right, a landscape in autumn 1998, on a misty day in summer 2008 and in winter 2018.
Source: Timo Pukkala, University of Joensuu, Finland. |
As we beat our paths through the information jungle, only the sturdiest tools will survive. Some of the technologies and systems that seem full of promise now will turn out to be short-lived; but some will grow and develop, no doubt in unforeseen ways, and provide greatly enhanced facilities to improve people's understanding and management of forest resources. Whether such facilities will be used successfully for the benefit of the whole world and all its peoples is another matter - humanity is more complex than trees!
References
Cisco Systems Inc. 2000. GPRS White Paper. Internet document: www.cisco.com/warp/public/cc/so/neso/gprs/gprs_wp.htm (posted 6 July 2000)
d'Orville, H. 2000. Towards a knowledge-based economy: the role of information technologies. Paper presented at the UNITAR Workshop on Information and Communication Technologies for the High-level Segment of ECOSOC, New York, February 2000.
FAO. 2000. First Consultation on Agricultural Information Management. Internet document: www.fao.org/coaim/
GIIC. 2000. Analysis of the digital divide. Global Information Infrastructure Commission (GIIC) Internet document: www.giic.org/images/Digital%20Divide%20 (GIIC)/index.htm (posted October 2000)
IUFRO. 1999. Task Force: Global Forest Information Service. Internet document: www.iufro.boku.ac.at/iufro/taskforce/hptfgfis.htm
Jensen, M. 2000. African Internet status, September 2000. Internet document: www3.sn.apc.org/africa/afstat.htm
Szaro, R.C., Martin, M.R. & Landis, E. 1999. Mobilizing scientific information on forests to promote their sustainable management in ACP countries. Paper presented at the GFIS side event, 3rd Session of the United Nations Intergovernmental Forum on Forests, Geneva, Switzerland, 3-14 May 1999.
UMTS Forum. 2000. Shaping the mobile multimedia future - an extended vision from the UMTS Forum. Report No. 10. Internet document: www.umts-forum.org/reports/report10.pdf
UNDP. 1999. World Development Report, 1999. New York, United Nations Development Programme (UNDP).
UNDP. 2000. Human Development Report, 2000. Internet document: www.undp.org/hdro/index2.html
