Cartographical Data and Data Quality Issues

This paper reviews three readily available cartographical databases (DBs), the Digital Chart of the World (DCW), the World Vector Shoreline (WVS) and 30 Arc-Second DCW Digital Elevation Models (DEM) all originating from the US Defense Mapping Agency (DMA). These are presumed strong candidates as cartographical data sources for strategic needs at several Consultative Group of International Agricultural Research (CGIAR) centres. While most cartographical themes found within these DBs are of acceptable quality, some themes definitively needs improvement. No known substitutes to these are known to the author. An approach is being suggested to overcome these inadequacies. Furthermore, the data quality part of the US Spatial Data Transfer Standard (SDTS) is being briefly summarised with the aim to stress the importance of data quality reporting within the UNEP/CGIAR project when data sets are created. INTRODUCTION, DEFINITIONS AND SCOPE I will in this presentation briefly summarise information on some GIS databases (DBs) that originates from the US Defense Mapping Agency, are currently available, are potentially suited for strategic needs of several CGIAR centres and have been recommended to be used as such at the first CGIAR/UNEP workshop (Arendal I). Strategic needs was at Arendal I defined as requirements related to studies covering a large geographical area, for example in characterisation and classification to be used in strategic research planning (anon. 1992). I will limit this presentation to cartographical DBs. By cartographical DB is meant a structured collection of digital GIS datasets digitised from topographical (or equivalent) paper maps or created for the purpose of making topographic type paper maps. Commonly, the following themes are included in cartographical DBs: (i) Coastline, (ii) International boundaries, (iii) National administrative boundaries, (iv) Transport infrastructure, (v) Cities, (vi) Hypsography or Digital Elevation Models and (vii) Hydrography. In this presentation International and Subnational Boundaries are excluded. These will be presented by Deichman and Fox later. Besides the value of this kind of GIS data for cartographical purposes, they are also of great value in GIS modelling. The use of administrative boundaries to derive statistics is a most prominent example. I will not examine sources for GIS cartographical data for operational needs. Operational needs was defined at Arendal I to be related to much more detailed data requirements for studies applied to relatively small areas. The distinction between strategic and operational data are here set at a scale of 1:1,000,000 (or resolution 1 km). Furthermore, I will briefly present the data quality part in the US Spatial Data Transfer Standard (SDTS). The SDTS is one out of several standards for geographical data (and information) that recently has been developed or are under development. The intent by