Tools for Integrating Geospatial Statistical Analysis into Spatial Data Infrastructures

Dr. Ben Gräler presents “Tools for Integrating Geospatial Statistical Analysis into Spatial Data Infrastructures” in Session III – Challenges and solutions for creating Geospatial Statistical Outputs. The joint UNECE/UN-GGIM Workshop on Integrating Geospatial and Statistical Standards takes place November 6 – 8, 2017 in Stockholm, Sweden. It is part of the Conference of European Statisticians’ work program for 2017. The objective is to highlight how geospatial data sources, methods, and standards can be integrated into the production of official statistics, and to identify a plan for future collaboration to support integration and interoperability of statistical and geospatial standards.

Abstract: “Statistical analysis is performed with dedicated tools and software. A prominent example is the R software for statistical computing. We outline an approach for integrating statistical analysis implemented in R into existing spatial data infrastructures using Open Source tools. Our approach is based on standards defined by the Open Geospatial Consortium (OGC), in particular the OGC Web Processing Service (WPS). The WPS offers an interoperable web service interface for describing and executing processes that transform some input to some output data (e.g. environmental models, GIS operators, etc.). In our approach, users can upload annotated R scripts to web processing services enabling their re-use by any interested party. The R packages SOS4R and sensorweb4R are used to flexibly assess spatio-temporal data from Spatial Data Infrastructures in R. The Uncertainty Model Language (UncertML) is used to integrate uncertainty measures in the results of the statistical processing. The usage of the software tools is illustrated by an application running at the Belgian Interregional Environment Agency.”