Session 3 : Integration of geography in the statistical production process

Geostat 2: Building and maintaining a point-based geospatial reference framework for statistics – preliminary conclusions from the project
Jerker Moström (Statistics Sweden, Sweden)

GEOSTAT 2 is a two year ESSnet grant project launched in February 2015. The aim of the project is to foster a better integration of statistics and geospatial information in order for the statistical community to provide more qualified descriptions of society, economy and environment in a spatial context.
One of its main objectives is to propose a generic model for a national point-based geospatial reference framework for statistics, building on national address, buildings and/or dwelling registers, suitable for statistics in the widest possible sense.
But what is a point-based geospatial reference framework and what are the challenges and benefits of using it ? Is it even possible to propose a generic model considering the great variety of practices among statistical institutions throughout Europe ?
The presentation will give a brief overview of the preliminary findings of the project so far. It will outline the concept of a point-based foundation for statistics and some of the operational aspects regarding setup and maintenance of such an infrastructure. Finally, the presentation will also briefly outline how the results from the project should be understood in relation to the Global Statistical Geospatial Framework.

Geostat 2 : Some experiences and conclusions - applying GSBPM to geospatial statistical data production
Rina Tammisto, Marja Tammilehto-Luode, Tuuli Pihlajamaa (Statistics Finland, Finland)

A Generic Statistical Business Process Model (GSBPM) has been introduced in statistical agencies for describing statistical production processes. In the Geostat 2 project, the GSBPM was applied to the geospatial statistical production process. A coarse level model was created to understand the GSBPM and the relations between ordinary statistics and geospatial statistics production processes.
In order to fit the GSBPM to cover geospatial production processes, quite a lot of interpretations of current guidelines are required. However, Statistics Finland found the interpretations possible. Even on a very detailed level, our interpretation covered the whole production related to certain single geospatial statistics. The challenge was, however, to transmit one interpretation to a generic interpretation, which could even be utilised by another organisation. To collect examples of using GSBPM in geospatial statistics production processes, Statistics Finland provides a ten point instruction on how we made our own tests.
In our presentation we discuss the lessons we learnt using GSBPM for describing our geospatial statistics production processes. We also argue that GSBPM with its phases should stay as it is but it’s interpretation should be extended by adding guidelines for geospatial statistics dimensions.  If we start, in this situation, to point out and stress the specialty of geospatial data (by adding new phases), we may lose a unique chance to consider spatiality as one dimension of statistical data and its important role in the general statistical production process.

Working towards a Geospatial Strategy in Statistics Denmark
Karen Skjelbo (Statistics Denmark, Denmark)

In Statistics Denmark we believe, that we have an unexploited potential for making use of the geospatial dimension in the production and communication of statistics. Therefore, our Board of Directors have established a task force, with the purpose of developing a strategy for making more and better use of Geodata and GIS in Statistics Denmark.
In Denmark there are good digital geographic data, administrative registers with full coverage and good connections and integration between the different registers and the law of Statistics Denmark provides the legal base for Statistics Denmark to get hold of all this data. The basis for increased use of Geodata and GIS in Statistic Denmark is in place.
The new task force consists of staff in Statistics Denmark working with GIS for analysis, GIS for visualization and staff having knowledge about geographical data. To ensure a broad approach in the work with the strategy, the leader of the task force is from the division of “Methodology and Analysis”. The work with the strategy is divided in to three tasks.
1. A phased strategy for an increased use of Geodata and GIS in Statistics Denmark. The strategy must contain principles for use e.g. standards, software, data modelling.
2. Specific proposals for new and better use of geography in statistics and analyses.
3. Benefits and costs for the realisation of the strategy.
A part of the work with the strategy is to get input and feedback from users in Statistics Denmark. But it is also considered to invite external experts and users of statistical data and private GIS companies to get inspiration for use and visualization.
The task force is just about to commence its’ work. In November the task force will be almost finished with task 1, and I will in my presentation describe the work process and introduce the first draft of the strategy.

Combining statistical and geospatial data – challenges and possible solutions from German official statistics’ perspective
Dr. Susanne Schnorr-Baecker, Malte Etienne (Federal Statistical Office of Germany, Germany)

Statistical data consist of a temporal and spatial dimension. This means that they are valid for a certain point in time or period of time as well as for a certain location. At present, official statistics already provides a wide-range of statistical data about social, economic and environmental issues which allows a deep insight in current developments as well as changes over time. Increasingly place based analyses for various units – administrative and non-administrative ones – at subnational matter in politics, economy and science. As a consequence, official statistics is asked to provide interactive maps with statistical indicators which are also included into comprehensive geospatial data-infrastructures (GIS) at national and international level for instance INSPIRE at European level. Additionally, new data sources, techniques and methods in official statistics allow new statistics and statistical information and new forms of collaboration.
Among cartographical mapping for administrative and non-administrative districts, grid-based mapping is becoming more and more important. However, the Federal Statistical Office of Germany had to learn from both the agricultural census 2010 and the population census 2011, which are also available on the internet how challenging grid based data presentation is.
Small area data based on grid-cells increases the range of information of Official Statistics and provides flexible and additional analysis options for users. However, the integration of geography in all phases of the statistical production process also goes along with some challenges to be dealt with of which confidentiality is clearly of major importance. Besides preserving the fundamental legal restrictions which are in Germany predominantly defined by the Federal Statistical Law, it is necessary for dissemination to address the usability of the outcome as well as the data quality. In this regard, considering the criteria and standards defined by European Statistics - the Code of Practice - helps to deal with challenges as it points out which core aspects must be complied with within the production process. This includes – just to mention a few – reliability, accuracy, timeliness and timely, the usage of appropriate methods or the obligation to operate in a cost effective way.
Some major requirements when linking statistical and geospatial data will be described in more depth during the presentation. Furthermore, possible solutions how the respective challenges can be mastered from the German official statistics perspective will be illustrated.