Inter-Agency Committee on Geomatics Canadian Geospatial Data Infrastructure Report Workshop ##1 Framework Data Held in Ottawa,  May 12, 1997 Prepared by Timothy V. Evangelatos Coordinator  Canadian Geospatial Data Infrastructure (EvangelatosT@dfo-mpo.gc.ca)

Quick Links 1. Participants: 2. Introduction: 3. Presentations: 4. Discussions: 5. Conclusions, Recommendations and Actions: 6. References: 7. Attachments:

1. Participants:

Norm Beck Natural Resources Canada 
Lloyd Bowler Natural Resources Canada 
Ed Buchinski TBS 
Mark Corey Natural Resources Canada 
Denis DeGagne Natural Resources Canada 
Caroline Erickson Natural Resources Canada 
Tim Evangelatos Inter-Agency Committee on Geomatics (Chair) 
Francois Faucher Natural Resources Canada 
Gordon Garrard Natural Resources Canada 
Arthur Goldsmith Environment Canada 
Ron Graves Statistics Canada 
Heather Kinsley Canada Post
Jeff Labonte National Defence/Natural Resources Canada 
Bob O'Neil Natural Resources Canada 
Andy Rencz Natural Resources Canada 
Peter Schut Agriculture & Agri-food Canada 
Dale Swam Environment Canada 
Danny Wall Statistics Canada 
Grace D. Welch University of Ottawa

TOP

2. Introduction:

This workshop was organized in response to a new trust of the Inter-Agency Committee on Geomatics (IACG), to create a Canadian Geospatial Data Infrastructure (CGDI). Several nations are currently working on creating national geospatial data infrastructures and the growing requirements for such data in Canada makes it an opportune time for a Canadian initiative. Some of the other national efforts are summarized in a study conducted for the Canadian Council of Geomatics (CCOG), by Geoplan Consultants (Reference 1), which recommended the need for a uniquely Canadian approach and suggested a cooperative approach in creating a CGDI. The report also suggested the provision of framework datasets containing core geospatial features supported by data standards defining content, structure and classification. Both the CCOG and the IACG have accepted these particular recommendations. 

To provide a context for the workshop Mark Corey gave an overview of related activities in the United Sates, Australia and New Zealand. More information on these activities can be found at the web site: http://www.fgdc.gov/ and in Reference 2. He then described the model, shown in Figure 1 that the IACG has recommended for creating the Canadian Geospatial Data Infrastructure. The purpose of this workshop was to define the data that should be included as "Framework Data" and to develop a strategy to make this data available. 
 

CGDI Model
Access to Data	Framework Data	Geospatial Standards	Partnerships	Supportive Policy Environment
Foster access to meta-data, images, files, and database query and extraction	Core data for control, transportation, topography, toponymy, elevation, and boundaries	Foster harmonization and convergence for collection, description, quality, access and dissemination of geospatial data	Foster cooperation to collect, build, share, and maintain the geospatial data	Foster access, lower cost, common licensing, and other activities to facilitate wider use of geospatial data

Figure 1

TOP

3. Presentations:

To provide the basis for the workshop discussions two presentations were given. The first presentation was a joint one by Norm Beck and Caroline Erickson of the Geodetic Survey Division of Geomatics Canada. Their presentation entitled: Geodetic Control: "A Foundation for a Framework?" gave an excellent overview of what is happening in Canada with regard to spatial referencing and provided a short tutorial on geodetic datums and reference frames. They concluded by describing the Canadian Spatial Reference System (CSRS), and how it will meet the needs of the developing CGDI. The slides shown in the presentation are included as Attachment 3.

Jeff Labonte then provided a perspective of foundation and framework data based on the efforts of the Department of National Defence, and the National Imagery and mapping Agency (NIMA). These organizations have been very active in defining and building a Global Geospatial Information Infrastructure (GGII).The presentation provided a conceptual framework for the foundation/framework model, described the constituent characteristics of framework data and detailed several of the experiences from the military environment.  Some discussion on the causations of this experience was also made. His slides are included as attachment 4.

The third presentation was given by Peter Schut of Agriculture and Agri-Food Canada who, following the discussion at the last IACG Meeting on March 18th, 1997, had submitted a proposal on the defining principles of a geospatial data framework and its components, which was used by Peter in his presentation. The proposal had been circulated with the agenda (Attachment 1) prior to the meeting. According to Peter:

The CGDI framework must: 
1. accommodate raster and vector datasets 
2. be upgradable (i.e. supports date stamping) 
3. be quick and easy to implement 

The CGDI framework should:
1. be lightweight 
2. be within the federal government's purview
3. allow for phased implementation 
4. allow contributions/cooperation/participation by provinces, municipalities, private corporations 
5. support practically almost all GIS applications, including all different scales (cm - km) 

Peter stepped through the his proposed components of a CGDI framework, shown in Table 1, and set the stage for a lively discussion, which is summarized in the next section, and led fairly quickly, to a consensus for a vision and a proposed implementation of a core framework data set.
Framework components
 

Component	Application
	Vector					Raster
Secondary	roads	hydrography	political	ecological	topography	Cloud-free LandSat
Primary	Major road intersections
Underlying  (hidden)	Geodetic Survey of Canada control points

Table 1

TOP

4. Discussions:

Although Peter Schut's proposal got very broad support, it generated a lot of discussion and finally led to decisions, although similar to those suggested by Peter, but differing in a number of the details. The discussions highlighted some of the different viewpoints between data "producers" and "data users".

All agreed upon that the underlying data would be based upon the control points and reference systems established by the Geodetic Survey of Canada, but the proposal to include major road intersections, as an independent set of point features, at the next level of the framework dataset created the much debate.

A road network of Canada is included in the National Topographic Data Base (NTDB), created and maintained by the Canada Centre for Geomatics in Sherbrooke. Additional road information is collected and maintained by provincial and other agencies. The road features are defined by a network of lines and appropriate software can readily detect the intersections of any of these roads, and so Peter had to respond to the question - why should a reductant set of features, such as road intersections, be added to the data set?

Peter responded very well with a highly convincing answer, using examples of his own work, where agricultural data is integrated with base topographic and other data including remotely sensed imagery. Although all agreed to the concept and use of an underlying geodetic control system as the basic reference system, in practice Peter and his staff use major road intersections to integrate and "rubber-sheet" their data to optimize the fit with other data sets. Other users expressed similar requirements. Furthermore, there was a requirement by the topographic data users to be able to add additional attribute information to these road intersections, which would not be feasible with the virtual intersections contained within the road networks. It was felt that road intersections should not be used to collect data but would be used to relate existing datasets. 

After considerable debate, it was agreed that major road intersections should be part of the core data set. In areas where there are few if any roads if was recommended that major rivers and shoreline should be included for the same purpose. It was also felt desirable to include major names and boundaries, but as dissemination of such data may have restrictions, due to agreements between federal and provincial agencies, there may be some limitations on what can be included in the initial core data set.

The discussions on content of the framework data set concluded with essentially a vision of four levels of data as follows:

1. Underlying (hidden) data: Geodetic reference system
2. Primary or Core Data: Major Road Intersections and Major Transportation Networks, Major Rivers and Coastline, and Key Elevations  - in effect a sparse subset of the NTDB.
3. Secondary Data: Toponymy and Administrative Boundaries.
4. Future Additions: More Detailed Elevation Data, Orthophotos and Satellite Imagery.

On behalf of Geomatics Canada, Mark Corey volunteered to take the initiative in producing a core data set and making it freely available, but he would welcome participation of other federal and provincial agencies in adding to this core data set either for the prototype and/or for future versions. Mark stressed the importance of keeping the core dataset very simple and easier to use.

To enable data to be made available at little or no cost and without copyright restrictions it was agreed that this should be a sparse subset of datasets such as the NTDB. 

The final discussion concerned the accuracy and precision of the data. Even though it was agreed that a sparse data set should be compiled it was felt that this data should be accurate and with as high a precision as possible, but 1/50,000 data from the NTDB was seen as practical and doable in a reasonably short time. This does not preclude incorporating more accurate data from the provinces or other agencies should it become available.

A lot of interest was shown in adding some form of imagery to the core data set, but it was decided to continue discussion on this type of data later.

TOP

5. Conclusions, Recommendations and Actions:

Although full support was given to the concept, and the overall vision of the core dataset, it is likely that participants left the workshop with differing interpretations of the exact details of the core data set. It is necessary to refine and document them, and bring all participants to the same level of understanding. The differences may be easier to reconcile, if we can keep the "vision" of the core data set separate from the "actions" that will be taken to achieve either part, or all of the proposed vision. 

Geomatics Canada has volunteered to provide a sparse selection of their NTDB data and a more detailed specification of this data needs to be prepared and circulated to the workshop participants as soon as possible This leads to the following proposed actions:

Action 1: Tim Evangelatos to work with NRCan to develop a proposed set of detailed specifications for the core data set, and to circulate these specifications for comment. If required, to also organize a second workshop to review and finalize the detailed specifications and to consider other related issues involved in providing the framework data. This would include defining what attribution would be included with road and river intersections.

Action 2: Tim Evangelatos to work with NRCan and other participating agencies to develop a plan for the construction of the core data set.

Figure 2

TOP

6. References:

1: "The Development of an Integrated Canadian Spatial Data Model and Implementation Concept." Prepared for the Canadian Council on Geomatics (CCOG) by Geoplan Consultants Inc. Fredericton, November 1996.
2. "Proceeding of the Conference on Emerging Global Spatial Data Infrastructure." Bonn, September 5th and 6th, 1996. The conference was organized by the European Umbrella Organization for Geographical Information (EUROGI).
3. A Project Proposal for a Spatial Modelling Framework for the Canadian Forest Service. Private communication from Jim Farrell, Canadian Forest Service, Natural Resources Canada.

TOP

7. Attachments:

1. Agenda
2. Framework Data Proposal - Peter Schut, Agriculture & Agri-food Canada
3. Geodetic Control: A Foundation for a Framework - Presentations by Caroline Erickson and Norrn Beck, Geodetic Survey Division, Geomatics Canada.
4. Global Geospatial Information Infrastructure (GGII) - Prensentation by Jeff Labonte, Department of National Defence.

Attachment 1

Canadian Geospatial Data Infrastructure
IACG Workshop
On
Framework Data
May 12
Room 204, 615 Booth St.

Proposed Agenda

09:00 May 12, 1997

1. Introduction and Goals      Tim Evangelatos (Chair)
2. Geodetic Control:
"A Foundation for a Framework?"    Caroline Erickson (GSD)
Nrom Beck (GSD)
3. Overviews of Other Framework Initiatives:
Global Geospatial Information Infrastructure (NIMA/DND) Jeff Labonte (DND)
National Spatial Data Infrastructure (US & Australia)  Tim Evangelatos
4. Towards a Canadian Geospatial Data Infrastructure Framework Peter Schut (AC)
Note attached proposal by Peter Schut, and also available from: http://res.agr.ca/CANSIS/NSDB/zcgdi.html. It is suggested that we use this as the basis of initial discussion.
5. Discussion:

Lunch:

6. Conclusions:

1. Recommendations to the IACG - Data Framework for the CGDI
2. Steps required to implement/complete this framework
3. Proposed work plan

(TVE April 24, 1997)

(Attachment 2)
Towards a Canadian Geospatial Data Infrastructure Framework

Overview
This proposal lays out a framework for a Canadian Geospatial Data Infrastructure, based on the discussions at the IACG meeting of March 17, 1997. 

Assumptions
The CGDI framework must: 

4. accommodate raster and vector datasets 
5. be upgradable (i.e. supports date stamping) 
6. be quick and easy to implement 
The CGDI framework should: 

5. be lightweight 
6. be within the federal government's purview 
7. allow for phased implementation 
8. allow contributions/cooperation/participation by provinces, municipalities, private corporations 
9. support practically almost all GIS applications, including all different scales (cm - km)

Framework components
 

Component	Application
	Vector					Raster
Secondary	roads	hydrography	political	ecological	topography	Cloud-free LandSat
Primary	Major road intersections
Underlying  (hidden)	Geodetic Survey of Canada control points

Status
 

Component	Source	Status	Actions required
			produce	align	publish	Distribute
survey control points	Geodetic survey	available now			X	X
Major Road Intersections	CGDI	produce ASAP	X		X	X
road network	NTDB 1:50K	partially complete	X	X	X	X
hydrography	NTDB 1:50K	partially complete	X	X	X	X
political  (provinces, CDs, CSDs)	1996 Census 1:50K	complete		X	X	X
ecological  (ecodistricts, regions, zones)	1995 Ecostrat	complete		X	X	X
topography	DTED?	?	?	X	X	X
Cloud-free LandSat	summer 1996	available	X	X	X	X

Issues
1. Should secondary components even be part of the framework? After all, agencies can release datasets that comply with the framework, and use the framework as a selling point whether the components are part of the framework or not. 

Finer details
For the "major road intersections" component: 

1. points should be identified by their NTS mapsheet and vintage (e.g 31G05A-1997). 
2. points should be identified on large scale maps, and on NTS 1:50K & 1:250K scale maps. 
3. there should be at least 1 point per NTS 1:50K mapsheet. 
4. where no road intersections exist, major hydrographic features should be substituted. 
5. points could eventually be permanently identified using 2 embedded geodetic monuments at opposite corners of intersections. (i.e. the point lies halfway between). This would permit accurate GPS triangulations, and supplement existing municipal control point networks. 
6. points should be identified to 1 cm. accuracy. 

1. the vintage of the primary framework should be identified 
2. generalized versions for 1:1M and 1:7.5M should be produced and made available.