Efficient Geoprocessing

With the development of novel and cost-efficient observation technology, the amount of spatio-temporal data about our environment is rapidly increasing. Due to the amount of data available and its heterogeneity regarding format, metadata, storage, and interfaces, geoprocessing tools offered in traditional desktop GIS are unsuitable and novel approaches for distributed geoprocessing are needed.

The 52°North Geoprocessing Open Lab addresses these issues and develops novel solutions for sharing, distributing, and integrating geoprocessing tools in different technical environments. These tools may range from simple operators, e.g. buffering line segments, up to complex environmental models relying on spatial information, e.g. flood prediction models utilizing in-situ and satellite observations.

A core focus of the lab is the integration of geoprocessing facilities in spatial data infrastructures by means of the OGC Web Processing Service (WPS) standard. 52°North has led the standardization process of the OGC WPS version 2.0 and provides a Java-based implementation, the 52°North Web Processing Service, which features a pluggable architecture for processes and data encodings and comes with several ready-to-use process repositories.

Assuming that geoprocessing tools can be easily shared and deployed flexibly  at various locations in spatial data infrastructures, the problem of how to find and execute these tools in a user-friendly way remains unsolved. Therefore, we develop novel concepts and clients for the discovery, orchestration and execution of geoprocessing tools.

Processing large amounts of spatio-temporal data requires pre-processing and post-processing steps. Novel approaches are also necessary for deploying geoprocessing tools close to the data instead of transferring data to locations running geoprocessing tools. Hence, another challenge is to identify necessary pre- and post-processing steps for big datasets and develop concepts for comparing and optimizing architectures for distributed geoprocessing.

The challenges mentioned above are largely technical challenges. We also deal with geostatistical issues, e.g. modeling spatio-temporal dependencies or extreme events and work on describing how data was gathered and processed in order to infer a meaningful application of models and operators to datasets.