Geospatial data users increasingly face the need to assess how datasets fit an intended use. However, information describing data quality is typically difficult to access and understand. Therefore, data quality is often neglected by users, leading to risks of misuse. Understanding data quality is a complex task that may involve thousands of partially related metadata. For complex cases where heterogeneous datasets have to be integrated, there is a need for tools supporting data quality analysis. This paper presents the design of such a tool that can manage heterogeneous data quality information and provide functions to support expert users in the assessment of the fitness for use of a given dataset. Combining concepts from GIS and Business Intelligence, this approach provides interactive, multi‐granularity and context‐sensitive spatial data quality indicators that help experts to build and justify their opinions. A prototype called the Multidimensional User Manual is presented to illustrate this approach.

Looking back at the last four decades, the technologies that have been developed for Earth observation and mapping can shed a light on the technologies that are trending today and on their challenges. Forty years ago, the first digital pictures decided the fate of remote sensing, photogrammetric engineering, GIS, or, for short: of geomatics. This sudden wave of volumes of data triggered the research in fields that Big Data is plowing today: this paper will examine this transition. First, a rapid survey of the technology through the succession of selected terms, will help identify two main periods in the last four decades. Spatial information appears in 1970 with the preparation of Landsat, and Big Data appears in 2010. The method for exploring geomatics’ contribution to Big Data, is to examine each of the “Vs” that are used today to characterize the latter: volume, velocity, variety, visualization, value, veracity, validity, and variability. Geomatics has been confronted to each of these facets during the period. The discussion compares the answers offered early by geomatics, with the situation in Big Data today. Over a very large range of issues, from signal processing to the semantics of information, geomatics has made contributions to many data models and algorithms. Big Data now enables geographic information to be disseminated much more widely, and to benefit from new information sources, expanding through the Internet of Things towards a future Digital Earth. Some of the lessons learned during the four decades of geomatics can also be lessons for Big Data today, and for the future of geomatics.

Guest EditorialInternational audienceThe need to combine data from different sources has become increasingly important and is likely to be so in our increasingly networked world. There are many difficulties that must be overcome to ensure the efficient and, to the es=xtent possible, the automated integration of different data sources. Although different goals and constraints may apply to the problem, not infrequently data sources are combined to achieve data of better quality. The process is generall called 'data fusion', other terms may also be applied to the same problem, including 'data integration' and 'data conflation'. For example, the term fusion is used by the image community to embrace the problem of sensor fusion, or by the database community for parts of the interoperability ^roblem. The logic community uses it for knowledge fusion.This special issue, which takes up a set of papers presented at an international workshop in April 2001 in Quebec City, addresses the problem of data fusion from such a wide perspective, that is, including a range of tributary problems and a range of disciplines. The workshop was supported by both European 5th framework projects and several Canadian GEOIDE projects