A scripting approach for integrating software packages and geoprocessing services into scientific workflows

Scientific workflow plays an important role in complex scientific analyses. Various components including local software packages and remote Web services can be integrated into scientific workflows. In the geospatial domain, geoprocessing workflows have been widely leveraged in spatial data analysis. Existing model builders, such as ESRI ModelBuilder and ERDAS Imagine Spatial Modeler are often limited to their own proprietary environments. With the development of Web Service technologies, an increasing amount of interoperable geoprocessing services are available. Workflow tools capable of integrating local and remote geoprocessing utilities are needed. This paper presents a scripting approach for integrating software packages and geoprocessing services into scientific workflows. The scripting approach is implemented by extending the GeoJModelBuilder, an existing workflow tool capable of supporting composition and execution of geoprocessing services following the Open Geospatial Consortium (OGC) standards. A hydrological analysis application of extracting drainage networks from Digital Elevation Model (DEM) data is used to demonstrate the approach.