Last updated July 29, 1996 by Lorant Czaran / 
June 17-20, 1996
PREFACE
INTRODUCTION
OPENING SESSION
DATA AND INFORMATION DISSEMINATION STRATEGIES
NARS, REGIONAL AND GLOBAL PERSPECTIVES
CENTER GIS IMPLEMENTATION STRATEGIES
After 1992 and 1995, the workshop in June 1996 was already the third time the same group convened to push forward the common interest of improving the conditions for the use of GIS in the International Agricultural Research Centers. This continuity and the fact that an increasing number of CG Centers is represented year after year - this time 13 out of 16 Centers sent at least one representative - is a sign for the popularity of the project.
The primary purpose of the Arendal III workshop was to establish the overall project direction for the next two years. The programme had been established with the following specific outputs in mind:
- Production of proposals for sub-projects relating to institutional development, data set production and/or application of GIS technology,
- Establishment of methods and procedures for data dissemination and maintenance activities among CG Centers,
- Defining and planning assistance to Centers to ensure continued productive use of GIS technology in the long-term.
Gwynneth Martin, July 1996
Chair: Svein Tveitdal
Rapporteur: Gwynneth Martin
Svein Tveitdal welcomed participants and gave a brief historical background, first of GRID-Arendal and then of the CGIAR-UNEP project. GRID-Arendal was opened in 1989 and is one of 12 GRID Centres world-wide, with specific areas of interest being Norway, the Polar regions, Nordic/Baltic regions, Central and East Europe, and developing countries.
In 1992, GRID-Arendal and the Norwegian Agricultural University hosted a meeting ("Arendal I") the recommendations from which were the basis of this cooperative project. The project received funding for the first year from the Norwegian Government and the World Bank, and the former has recently confirmed a commitment of a further $550,000 to the project over the next two years. In making this commitment the donor stressed the need to ensure that the CGIAR/UNEP-GRID network be self-sustaining at the conclusion of this second phase.
Given this background, the intent of this Workshop was primarily to establish the desired outcomes of the project's second phase. ST proposed the following as a starting point for discussion:
Claudia Heberlein reviewed the workshop programme (Appendix I).
It seems clear that there are expectations that new tools are needed to deliver technologies at the local level and that this project has potential to assist in meeting those expectations. The Workshop will be addressing what is wanted in the short-to-medium term, the role of GRID-Arendal, and how the project activities might fit into the CG research agenda in the longer-term.
1. Management and Networking
- Monthly electronic newsletter
- Brochure
- Arendal II workshop report
- Project proposal 1996-1998
- Participation Centers Week 1995 and Impact Assessment and Evaluation workshop
2. Institutional Development
- ToR Requirement Studies
- Requirement Studies in 10 Centers
3. Dataset Production, Data Dissemination
- Biodiversity Map Library
- Asia Population Database
- Administrative boundaries Latin America
- (planned) Climate Data Asia (CRES - ACIAR)
- Data Catalogue
- WWW
Further detail was given in two major areas.
It was noted that the Asian Population Database & Administrative Boundaries of Latin America will be reported on in a presentation scheduled in the "Applications" session.
The following points arose in the subsequent general discussion on the outcomes of the workshop and future direction of the project.
Rapporteur: Lorant Czaran
The session began with some introductory remarks from the Chair on the data exchange possibilities offered today by the INTERNET.
Both A paper version of the catalogue and a digital version (Excel spreadsheet) were distributed.
Two immediate points were made by participants:
- the need to link this to other homepages and related information
- the requirement to ensure the information was used otherwise the compilation is wasted effort.
Early versions were distributed to National agencies, with positive reactions from some (Kenya and Madagascar) and questions raised by others (Uganda).
Even Husby demonstrated the HKT. The software was later in the week handed out to all the participants on floppy disk.
In September 1995, ISNAR had organised a workshop for Centers on Web use, including development of homepages. At present, the situation is confusing, as several Centers now have pages at the same time as both CGIAR the World Bank maintain independent as well. There is a urgent demand for streamlining and linking.
ML mentioned several related activities:
- there will be a workshop on database publishing on the web in July at ISNAR
- applications may be established on the Web e.g. the financial system of the CG Secretariat
- IDRC has donated US$100,000 to provide databases through the Web.
Even Husby gave a second live demonstration, this time showing how GRID-Arendal is using WWW, and he outlined plans for the future. The demo pages can be accessed on http://www.grida.no/~husby/webtalk.
Vincent Abreu described the Land Quality Indicators (LQI) Project, recently begun under World Bank sponsorship. A major activity is the establishment of a global metadatabase system for LQI related information. Inputs are expected from FAO, UN Statistics Division, the CGIAR network, GRID Centres, GIS producers or users, UNEP, and they will include descriptions of tabular and other datasets. The metadata may be distributed and locations may each have different standards but there will be common fields. In addition to this metadata, interactive access to the actual data is also planned. Information on the project can be found through the CIESIN homepage.
A prototype is expected to be completed by August '96, for beta testing by users. Release of the first version is planned for December '96 and the fully functional version will be launched in June 1997.
- join in the on-going UNEP/GRID metadata activities
- make metadata information available to a coordinating unit (GRID)
- follow the recommendations made by GRID for information exchange.
Group 1
Group 1 reported agreement on the overall proposal, especially the usefulness of, and need for, a metadatabase . All Centers should join the effort to update and maintain this database. The metadata might also include information about future development and plans for datasets as this may be of interest, particularly to donors.
The group also agreed that UNEP should co-ordinate the collection of data from Centers and update the system on a regular basis. Centers need to make their contribution to this and must make a firm commitment. However, the HKT should be more flexible to include different types of data, and Centers should be able to use it without difficulties, in every location for their own purposes. Standardising the format of data input is also accepted as being important, and UNEP should follow up on this matter.
Quality control is primarily to be done by Centers themselves, as resources are available. It is expected that UNEP may develop some tools for assessing quality of data in the future.
Dissemination of data may be done by UNEP, but Centers may also be able to do this, as resources permit. Part of the group considered that it should be the responsibility of Centers to decide on dissemination, media, prices and accessibility. Various methods and forms of distribution (CD's, through NGO's, etc.) should be considered.
Group 2
The conclusions of Group 2 were on the whole in close agreement with those of Group 1.
Although a possible role for CGNET in co-ordinating data collection and update of metadata was considered, it was concluded that UNEP has more capability to do this, and it should undertake this activity for at least the next two years. Centers should follow recommendations from the GRID Centres regarding the use and production of the DIF files, and for the interchange of data and metadata available.
The question of socio-economic data was problematical, mainly because of the low resolution of the data sets available.
Schedules and milestones should be identified, with data to be provided by Centers as soon as possible. Dissemination methods should include alternatives to on-line solutions e.g. other types of magnetic media.
Group 3
Group 3 had concentrated their discussions on strategies and plans rather than any technical issues. The CG Centers can benefit from the UNEP-GRID work and should take advantage of that. Further development of the metadatabase (data catalogue) should be undertaken by GRID, utilising the funding available over the next two years as required. After that it would likely remain cost-effective to do it centrally, possibly CGNET could be funded through a small subscription from each Center.
CG Centers have a need to improve their data management practices. The use of the GRID/UNEP HKT should be encouraged if it is flexible enough to meet those needs. This requires evaluation. A suggested approach is to form a working group (coordinated through GRID-Arendal) with the following workplan:
- carry out tests of HKT (through to Dec. 96)
- report and determine modifications required (April 97)
- deliver prototype of modified HKT (Oct. 97)
- establish operational system (Jan 98)
The need to involve outside organisations such as NARS to determine data and metadata requirements was stressed.
Use of the GRID HKT should be seriously investigated. It was repeated that Centers have internal requirements to manage their information and whatever tool was adopted should be capable of use by Centers themselves to meet their own needs.
The need to include indications of quality in metadata was discussed, and the requirement to have some quality control on datasets which are made available.
Groups had concentrated on metadatabase issues and had not had time to pursue discussion of dissemination of data themselves.
Rapporteur: Claudia Heberlein
The four main areas of activities of ICIMOD are:
1. Documentation and Information exchange
2. Research
3. Training
4. Advisory Services
The Center is regionally oriented serving the rural mountain areas of Pakistan, Nepal, Myanmar, India, China and Bangladesh. The use of GIS in these mountain areas presents considerable technological challenges due to the inherent diversity, marginality (most of the areas lie in border regions), and varying biophysical and socioeconomic values. However even more critical is institutional cooperation and developing coordinated strategies for this are essential.
Within the MENRIS (Mountain Environmental and Natural Resources' Information Service) network, the training component plays a major role. Training addresses policy makers, professionals and technicians, and is carried out through case studies and training courses. The training which initially was done by ICIMOD is now more and more taken over by the focal points in the member countries.
Examples of how the network functions were presented from Nepal and Bangladesh. The players include national planning commissions, Government agencies and universities.
Rapporteur: Nickolai Denisov
Policy and technological scenarios are developed for ca. 20 years ahead with the purpose of estimating the long-term implications of R&D. The (dis)advantages of cross-border collaboration is a subject of special interest.
ML outlined advantages and disadvantages associated with GIS and suggested the following questions needed to be answered in order to ensure the efficient application of GIS technology:
The evaluation proposed is essentially a cost-benefit analysis. Case studies would be selected and the following criteria were suggested:
There were several immediate questions from the audience. The general feeling appeared to be that it would be very difficult to carry out a positive evaluation due to the timeframe required to show real benefits, the difficulty of attaching value to GIS datasets, and the need for overall institutional adoption of the approach, which is again taking time.
Gorkha is one of Nepal's 75 districts, with small area and very varying topography. Large scale applications (1:50,000) with a 3-dimensional view component are essential.
Although a lot of information was available, it was not primarily in digital form. The data layers now compiled include baseline data, land use/ land resources, and settlement-related parameters (population, agricultural statistics, etc.). For the latter (1100 settlements), the georeferencing was not always easy and approximate location data were sometimes used.
The GIS applications include:
In response to a question, the budget used to assemble data was US$40,000, including a workshop and some other related activities.
The main problems with socio-economic data arise from inconsistencies with regard to georeferencing, geographical names, scales/ accuracy, and missing/ obsolete information.
Full detail is given in the 28 page paper distributed.
Still missing are data for Mexico. Uncertainties are associated with El Salvador/ Honduras border area and a few other similar areas. Gaps still exist for some third level units (e.g. Paraguay), due to difficulties in getting corresponding maps, and there are still some errors in the CA part.
There are ca. 10,000 urban centres in Latin America; the next challenge is to attach the population data, after the geo-registration of the urban centres is complete. The move to the 4th level is also technically possible though the potential resource requirements render it impractical.
Data collection was the most time-consuming component. The following data collection strategy was applied:
This made determination of data quality difficult so informal summary measures were used. A six-step approach was used to build the population density and some product evaluation carried out. Details are documented in the paper distributed to participants.
The output products/data will be disseminated as 2-5 minute resolution data-sets, also available through the WWW. The three main products will be a data-set of population counts by grid cells, a population density data-set, and corresponding TIFF image(s).
The next steps should ideally focus on:
The programme focus areas are land-cover/ land-use, water resources management, loss of biodiversity, and impact of climate change. The issues to be covered are capacity building/ training (with a focus on North-South imbalance), integrating research and operational communities activities, improving information for decision-making, and harmonisation/ standardisation of databases (no systematic effort has been made so far).
A pilot project on data integration, in a partnership with the government of Norway, involves 8 countries in Africa, Asia, Latin America and Europe. The purpose of this is the elaboration of a set of priority monitoring/ reporting parameters, the testing of data transfer/ exchange procedures, and the development of harmonisation/ standardisation approaches for data collection and dissemination.
A hierarchical model of indicators (poverty/ food security/ environment vs. institutional/ management/ technical level) was presented. The theory/ philosophy of modelling and the reasons to use map analysis in the process (pattern recognition, data filtering, alternative to mathematical languages) were discussed.
Rapporteurs: Claudia Heberlein, Gwynneth Martin
Proposals had been requested from Centers in advance of the Workshop and OS distributed a table summarising those in hand. This is not to be considered a closed list as other proposals may arise during the course of the workshop, or from follow-up activities. Those that are listed are in varying stages of completion -- several need additional information and most need clarification of some kind. The process proposed for compilation of the portfolio in the working sessions ahead was as follows:
- a standard outline for the proposals be adopted
- all existing and new proposals be completed following the standard
- a review of proposals submitted be carried out to ensure that each was relevant to the project, to group related activities if appropriate, and to prioritise them.
A strategy for approaching donors would also be developed at a later stage.
Each proposal was then outlined briefly by the person who submitted it.
As time ran out, there was no clear conclusions reached on how to amend the Workshop Programme to reflect the opinions which had been expressed. It was agreed that Svein Tveitdal would begin the next morning by putting forward a proposal as to how to proceed further.
However ST reiterated that participants had the role of "General Assembly" in setting the future direction of the project and, with the above as a basis, he recommended that Working Groups focus on the following four issues.
B) GIS consultants?
C) A combination - if so what is the most important? Other?
A) Business as usual: UNEP/GRID-Arendal continues involvement on a project basis
B) Self driven by CG-Centers: Ad hoc contact, or one Center appointed as lead institute
C) Institutionalization: Official CGIAR body with a permanent secretariat
D) Other solutions/no need
A) Criteria to be included
B) Guidelines/harmonization procedures
C) Donor approach
D) Role in implementation
A revised program was suggested to accommodate group work along these lines (Appendix II).
Participants signalled general agreement with the outlined suggestions. In response to some organisational questions, it was agreed that there would be three working groups (as previously) and each group would have the opportunity to discuss and report back on all issues. The revised programme was then accepted.
Rapporteur: Otto Simonett
ST then listed seven parameters which were proposed as "indicators" of GIS usage and facilities in Centers. Center representatives had completed a brief questionnaire (Appendix IV) based on these and a matrix had been built and was now presented. Clearly this is not a strict or complete methodology but was presented as a way of summarising the various Center approaches. It was stressed that the appropriate strategy was not necessarily to have a complete ("totally green") matrix.
The following comments were made:
- the role of third parties should be included, e.g. to indicate such things as WCMC's role with CIFOR and ICLARM
- the contracting-out parameter was ambiguous (the line was removed in response to this suggestion)
- an additional indicator was the (GIS) project work carried out in one year.
Copies of both the questionnaire and matrix were distributed and further input invited from participants.
CIMMYT has been assisted by the project in development of strategy and workplans.
The animated discussion revolved around issues of perceived (or real?) incompatibilities of funding database maintenance through projects. While some participants see increased pressure to fund their activities through projects as a threat to databases, others see opportunities. Germplasm databases have been put forward as a (contested) success story.
Another issue was cost-benefit considerations. Svein Tveitdal pointed to his studies in this area in the past and indicated that in 'hard' cost-benefit analyses of data development activities, social benefits are neglected. The GIS data domain can be regarded as an incomplete market. The audience showed considerable interest in these studies, and suggested that the reference material should be distributed.
The GIS programme of the institute is relatively new. The target groups identified are use of GIS by managers, by researchers, for communication, and for global decision-making. Activities focus on workshops, special 'dashboard' systems (for managers), support in spatial data, GIS modelling, platform for modelling and remote sensing (research). GIS at the local level are used to demonstrate the usefulness of the technology as a communication tool. This was presented at Arendal II last year. A global application mentioned was the creation of a global climatological atlas and database using a large number (28,000) of meteorological stations.
Rapporteur: Gwynneth Martin
- identifying available datasets,
- disseminating the data catalogue on Internet and CD-ROM, and
- obtaining data for CG Centers upon request.
The Group also felt that it was unclear what would be involved in institutional development activities, following up the requirement studies and that some plenary discussion of this was needed.
b) If funds are limited, the secretariat's catalytic role was seen to be the most important. However if there is any leeway at all, a consultant role is highly desirable e.g. to develop cost-benefit figures for GIS usage, to recommend standards, etc.
c) To ensure sustainability, the Group suggested the option of creating a System-Wide Initiative cutting across existing programs, following the model of the existing participatory methods initiative.
b) It was suggested that, as a result of the Center requirement studies, recommendations as to where assistance was needed might be made, and perhaps even specific names could be given identifying sources of the expertise needed.
c) In the long-term there was a desire to keep the Arendal initiative going and it was felt that the CG system should find the resources required. Options as to how this might be done were left open.
- identification of priority data needs, in detail, and
- provision of data management tools.
The data catalogue was felt to be important and the Group referred to the specific steps suggested in the previous Working session (section 2.7). The proposed package for awareness building must address costs and benefits of GIS. The use of a list server was suggested, possibly to replace the monthly newsletter.
b) Institutional development activities should look at the broad issues of information management.
c) In considering sustainability, the creation of an efficient market was seen as key, i.e. identification of the benefits realised from the sharing of datasets/expertise/networking was essential.
Svein Tveitdal recognised the need expressed for an awareness building package and saw it as a priority for project activities. Similarly, follow-up on Center requirement studies would be pursued although there was mixed opinion on the optimum timing (short- or medium- term).
Donor representatives were asked to comment on the question of sustainability. The ODA experience was that a network, once established and proven, became sustainable through a subscription process. Ruth Haug of NORAGRIC supported the pursuit of establishing a System Wide Initiative in the medium-term.
Proposals should:
- deal with strategic issues involving a number of Centers
- have a case-study component in 1-2 countries
- lead to outputs relating to research priorities.
Possible themes for inclusion in the portfolio were poverty, sustainability, environment and health, and marginal lands.
A primary criterion is that proposals must be of interest to/benefit to/endorsed by multiple Centers. In addition, they should produce results in the form of datasets created, methodologies established, or capacity building of a generic nature (involving strategic issues). Resolution of which proposals are to be included is likely to be best achieved through an iterative process.
GRID-Arendal's role in implementation was seen as the distribution point for results. They may have no other role but they may also be identified as an active partner in some proposals.
The Group made the point that the creation of the portfolio was not to be viewed exclusively as in pursuit of funds. It also served to coordinate Center efforts.
- generic nature (tools, data, methods of interest/applicability to more than one Center)
- potential for leveraging funds
- strong GIS component
- impact with NARS (though there was some concern that this might raise issue of competition with other initiatives)
The following process was suggested to produce a harmonized package:
- Centers prepare project briefs (2 months)
- briefs circulated for Center comments/evaluation (2 weeks)
- Steering Committee review (2 weeks)
- development of selected briefs into full proposals (2 weeks)
- portfolio harmonised (1 wk).
It was emphasised that Center management needed to be fully aware (and approve) of the contents of the package before any approach could be made to donors.
It was agreed that the Steering Committee should suggest the guidelines before any other decisions regarding the portfolio were taken.
The Working groups had given an extremely positive response on the usefulness of the project and had laid substantive groundwork for detailed planning. It is important to continue producing visible outputs of the project activities.
The discussions had not defined specific involvement for NARS and applications with national colleagues should be pursued. Similarly, relevant links need to be made with organisations outside the CG system who are also responsible for agricultural research activities.
Workshop participants ('The General Assembly') had provided guidance to the Steering Committee who were to meet next morning. In relation to the project portfolio the suggested criteria were very helpful. The Committee's task is to take those inputs and ideas to structure Phase II of the project, balancing the desired activities and available resources to ensure significant achievable results.
The first question from the floor was procedural, concerning how follow-up from this workshop would happen. Members of the Steering Committee responded by saying:
- a concrete workplan would be developed and circulated to participants, and
- the possibility of a System-Wide initiative would be pursued through CG channels, possibly to be an item for DGs/DDGs in ICW'96
Stan Wood volunteered that IFPRI would be interested in being involved in economic evaluations of GIS impact. Others echoed this interest.
With respect to data requirements, while specification of requirements might be undertaken, it should be recognized that this is a moving target, both with respect to content and resolution needed.
The question of the next workshop was raised. It was thought that a change of venue might be in order. CIAT offered to be a host, however, there seem to be practical problems related to US American security policy. Nairobi and Geneva were mentioned as alternatives. It was left for the Steering Committee to decide. It was suggested that the title 'Arendal IV' be retained regardless of the location as it symbolises continuity and has positive connotations.
Ken Fischer closed the session and workshop by expressing gratitude to those who had acted as chair and rapporteurs during the sessions, and thanking all the participants for their contributions.
Last updated July 29, 1996 by Lorant Czaran /