Augmenting geospatial data provenance through metadata tracking in geospatial service chaining

In a service-oriented environment, heterogeneous data from distributed data archiving centers and various geo-processing services are chained together dynamically to generate on-demand data products. Creating an executable service chain requires detailed specification of metadata for data sets and service instances. Using metadata tracking, semantics-enabled metadata are generated and propagated through a service chain. This metadata can be employed to validate a service chain, e.g. whether metadata preconditions on the input data of services can be satisfied. This paper explores how this metadata can be further exploited to augment geospatial data provenance, i.e., how a geospatial data product is derived. Provenance information is automatically captured during the metadata tracking process. Semantic Web technologies, including OWL and SPARQL, are used for representation and query of this provenance information. The approach can not only contribute to the automatic recording of geospatial data provenance, but also provide a more informed understanding of provenance information using Semantic Web technologies.