Document Sample
Attendees Powered By Docstoc

Attendees (Names are available in detail on attendees list: the ones below were done
verbally and therefore may contain errors):

Heinrich du Plessis: DLA CT, SA: Custodian
Anthony Cooper: CSIR built environment. Standards and curation
Kevin: Stats SA: interest
Louisa van der Merwe: CSIR Stellenbosch
Anna-Marie Arnold: CSIR library: need to know for preservation perspective
Ingrid Backgam: WRA funding: interest: own digital data curation within WRC
Alfonse Dubi University: Dar es Salaam: Research, teaching: spatial data interest
DST: Filicia?: interest
Pricilla: Mafikeng: GIS Consultant: research on how GIS to use in Africa millennium
Ande: AINA: interest in offshore maritime systems
Andre Roos: Nat Archives SA Dept arts and culture: preservation. Serves on sub-
committee for Geo-information standards. Special interest
Martie v Deventer: CSIR: start work with researchers on spatial data curation
Willem Coetzer: Aquatic biodiversity: GIS background: national fish collection:
Grahams town
Paul Uhlir: US: Archiving and policy for preserving scientific data
Heidi: custodians of digital data, esp. national landcover satellite data: how to keep


Heidi distributed documentation: OECD guidelines

Heidi provided an introduction to the workshop: she noted that the main focus should be
to find solutions for Africa on spatial data curation

She reviewed the OECD:
   - Principles: importance of cost-effective access to research data
   - Collective standards that member countries; but African countries are not
       members. S Africa is an observer, only African countries.
   - Q: Should African countries be part of OECD or will it have its own “internal”
       continental agreement”

DST National access workshop: commented on guidelines and suggested a plan of action
  - Data management is the scientist’s second priority only, not the principle aspect,
     i.e. if we give researchers incentives, they will curate their data better.
  - Danger of researchers reduced to gatherers of data that is analysed in Europe and
   - DST outcomes: challenges:
         o Education
         o System
         o Managerial
         o Institution barriers
         o Legal and policy issues
   DST decided was to:
         o Create a plan of Action
         o Do recommendations via parliament
         o Consult and assume full implementation of the OECD guidelines within

The main question that the workshop aimed to answer:
What are the recommendations for African countries to curate spatial data?

NeDIC is more “general’ outlines of systems: this is for all systems, but we now have to
look at specifics for spatial data in Africa

1) How could scientists / funding organisations be promoted to curate digitally produced
spatial data more responsibly?

2) How can government contribute to the creation of incentives?

3) Performance criteria / evaluation / acknowledgement / how to give credit and
incentives: is it a minimum requirement or a performance rating?

What’s the current situation in SA: Derrick Clarke (Anthony gave comment):
Study was commissioned by Surveys and Mapping, with partners across Africa: went to
26 African countries: what are the fundamental spatial data sets and how to get it all
available. Second report has just been published: The list of core data sets was reviewed
and Anthony discussed the useful documents: avail on EIS website (www.eis-

He noted that a lot of this data is available in repositories outside of Africa; and a lot of
data is still only available in hard copy maps. One problem with all this data was that they
were lacking in metadata. Best metadata came from foreign repositories. They produced
ISO 19115 Africa metadata standards: draft was produced. Thus: there was something
already done: there is a lot of data in Africa. It did not really address the curation of data:
there were a lot of issues e.g. bandwidth: e.g. landsat available to all of Africa, for all of
Africa: no requests received: people do not have funding to get CD’s, PC to read data;
and bandwidth to download. Dar Es Salaam requested it – he will contact Anthony

Spatial Data Infrastructure Act: and NSIF of DLA: Andre Roos: mentioned the role of
data custodians: NSIF plays coordinating role. Suggested that it should be implemented
for each country in Africa.
Suggestions on incentives: start with Policy per country > then get regulations, and then
get rules.

Who should implement it?: Research institutions: no / Funders should in the contract
provide the incentive. The OECD guidelines do not provide guidelines to funders -
since it dealt with all kinds of data. Specifics are left to the funders. Motivation can be
done top-down or bottom-up: Money, recognition, prizes, rewarding people o do the right
thing. More attention to be devoted to these issues at universities: will come though in
the next generation of researchers. Some universities are already doing this: probably
they are not addressing the curation. Most probably only acknowledging the data sets.
Bottom-up: motivating the researchers through mechanisms e.g. incorporation of the
value of the mechanism, in education and training, so that there is an appreciation for the
value of this type of thing.

Would this be a minimum requirement?: Can’t impose a step but don’t provide the

“Categories” of spatial data sets: research data: “product” and metadata with spatial data
set: larger the study area and more imagery used, the more space there would be needed.

Anthony noted: fundamental data sets that everyone uses: large custodians/orgs/govt (e.g.
aerial photography);
Project-specific / Research data sets: this is the tricky part: that needs to be archived.
Don’t always need the intermediate steps: someone should be able to reproduce the steps
and get the same results.

Mr Roos: Gov can indirectly urge people to contribute data for safety-purposes: should
be free of charge: that agency will make it accessible without metadata etc. Google
seems to be blundering into this domain. To keep repositories safe and help make the
datasets accessible; also by publicizing the existence of the data sets.

Would it not be the responsibility of the organizations to keep the digital repository?

Individuals also have the responsibility to curate data.

New attendee: standard citation form to be implemented: way to formally cite spatial data
that gives credit to the researchers.

Who would be responsible for creating the awareness: should be formally part of the
curriculum: formal inclusion in curriculum.

2: Solutions for known challenges:
What is the main problem(s) for Africa:
    - Spatial data in particular: most mapping agencies either fall under the military and
       fall under that background and fall under “classified info” (even though one can
       probably get more current and higher resolution images from Google), but it is
       used as an excuse for not sharing data.: Solution: Gov to change: legislative
       process. Also embarrassment process to be brought in: “if you don’t want to give
       your data we will get it from somewhere else”:
   -   Tremendous value of using such data for development and economical and social
       application: millennium development goals and linking it to the policy goals of
       each administration in each country.
   -   Whose job will this be to do it?: Ministerial champion: ministry of science and
       tech / some kind of combination: very important that it is clear what the goal is:
       may not necessarily need to be a particular dept. As long as there is a champion.
   -   Fear that it will be used by people outside the country to exploit local
       cultures/people. Especially with regard to mining data and conflicts in this
   -   There are developed countries that make their data available, and they have not
       had huge negative problems associated with it, in fact, they have benefited from
   -   Data collection: initially to be collected / collated by Gov and also Aid agency
       funded: Aid agency is sometimes ignored and reproduced, using different
       standards and technologies, then do their work and leave. And then there is
       commercial interest: Eg Eskom is doing very detailed surveys in Africa for power
       lines: they might or might not make it available, probably at a cost. Same with
       the cell phone companies.
   -   Last 3 on list was much discussed, what about the top 2?: Educational and
       systems: Dar Es salaam: it will need promotion at Govt level. Ministries of
       Science and tech and depts. Of education should address it, and be driven by both.
   -   Must be part of the curriculum; since it will then be automatically be part of the

1 year ago: call for nominations: lot of depts. Did not respond: first delay: then
submission to Minister was made: wanted more briefing: no-one driving that process at

CSIR is driving factor in SA: other organisations play a similar role in other countries.

What happened in NRF to the “SA data archives”: does it still exist? – this is their
conference: the name seem to have fallen away.

Who will benefit?: include research organisations and researchers, and org institutions
and students. (since these are sometimes different from those people who do the actual
spatial data work). Also add NGO’s (could be allies in promoting both the benefits from
these activities as well as making the case; and add the Citizens: (N for NGO’s to not
duplicate efforts).

Spatial data is a bit diff from other research fields, in that it is widely used, not for
research, e.g. also the UN ECA, UNEP and Pan-African Organisation and private
foundations, which are not funding organisations per se (and e.g. peace parks in SA).
Who are currently undertaking sufficient curation initiatives?: Rather call it “significant”:
SA Post office for address coding: spatial data sets?

“Curation” is a loaded word: “the data is kept somewhere” Curation may not be the
strength of a organization.

Curation is not the same as storing data on a PC: we do not have stunning role models,
since the data is not “available”, although it exists and is well managed. Every generator
of spatial info gatherer will store it: give them directives as to what their responsibility is:
SA has various acts, this may or may not be the case in other African countries.

Privately owned info is problematic: research may be kept with its value being added and
continue further high-level research is problematical and what regulatory systems could
be output into place to ensure that such data is available. Enable people to access info.
Should it be a similar structure such as spatial information?. What is available? Where is
it is available from? What are the standards.

Can take SA model and take it to the rest of Africa: maybe adapt for other counties.
Work through network of sub-organisation in African context. NEPAD is a good idea to
bring in.

Large industries generate huge amounts of spatial data: most is proprietary: opportunities
exist: at some point that data is not so proprietary any longer. E.g. seismological data is
no longer proprietary: either gets destroyed or “sits” on a shelf somewhere: need pro-
active mechanism for getting data that is no longer proprietary. Also regular reporting
from the industries to the Govt – some of that info, or at least a synthesis of the data, to
be available: still proprietary info, but can be used to generate maps.
Public-private partnership for infrastructure capability where the diff industries can
collect data on behalf of govt. e.g. meteorological office: get data: data not to be sold, but
made available freely.

Flying of same area and share costs, between organizations.

Citizen-created data: inexpensive technologies: have people collecting environmental
info: e.g. tracks for Africa: GPS data from people who have GPS’s in Africa: there will
be some niche applications where people will start doing what government agencies/big
companies are doing, themselves.

Biggest hindrance = coordination and implementation of the solutions.

Geospatial data is regulated in SA: belong to the state. Private sector data is
unregulated and the frustration is that research info is sitting in that sphere. Also
difficult because the data has commercial value.
Once it has no value any more they may even have destroyed the data, which makes the
data lost for ever. Private to private = problem.
Metadata availability = very important.

Solution ultimately multi-tier:
1) Per science domain: bottom-up approach

2) Government level: Give mandate to Ministry of S & T to take this need to other
African countries.

3) CSIR/Org level: Need regional cooperation between organisations (other African
Countries) may be even further developed that SA

4) Researchers: research criteria:

4) Funding agencies: place requirements into contracts and requirements.

5) Education: Add into curriculum

Centralised vs distributed responsibilities: important where it is done and how it is done:
institutions have their own standards: centralized is easier (e.g. national archives): up to
each country and even organisations: decentralized but inter-connected with common
standards, formats and federated data management.

Possibly set NSIF as standard for rest of African countries, to follow same principle: each
country should have a spatial data clearing house and standards generation body.

Non-network distribution of data: For a long time this will still be the case. May have
catalogue on-line, but CD’s get distributed regularly still: how long? – the volumes of
data is increasing all the time, but hard-distribution.

Who decides what data is worth archiving: science councils should be responsible for
regulating. Want an educational Act or amendment that state that the same obligations
are on educational institutions unless it is submitted: and centralized and archived in
suitable environment where it can be archived. And the understanding of “info”: people
don’t think of “spatial data” as info: subtle changes to existing acts, to also include spatial
data: in SA the principles are already in existence. Other African countries should then
follow suit.

Spatial data and geospatial information (e.g. imagery) will expand at e.g. 100 Terrabytes
per year: millions of rand to curate. Metadata is easier to manage. So that kind of digital
spatial data will always be de-centralised.

Also must limit it to “public” info. Funding org must specify whether it is public or not.
Money to implement legislation would not be beneficial either. Want to create a culture
of “this is the norm”, and do not necessarily want to regulate it everywhere.
Technological: non-proprietary formats and systems for long-term storage is a problem:
have problems with formats for storage and access. Particularly nb for imagery.
Standards = nb.

6. Way forward:
Champion needed: UN ECA can be a champion: has a committee for spatial info
NEPAD? – is it functioning? – does not have the subcommittee either.
Rather than a national approach maybe a disciplinary approach, e.g. oceanographic:
already have network, already doing data curation: bottom-up approach through
disciplines. All the people involved have a common understanding of the curation issues
within a specific research field.
Needs-approach: it is needed.
And might also have a quicker route into funding.

A discipline-specific approach is more likely to succeed in the short term: capacity is

The campaign should go on: eventually it will be taken up by African scientists: begin
with the scientists. Politicians will hear it but there may not be a lot happening from it.

Closer integration of different Acts, esp. in SA. Proposed amendments to Acts could have
an impact: question of terminology “record” /”document “data set”: need to be referenced
correctly in the new revised acts.