The Digital Earth Software Framework

The investigation of scientific research questions can be conducted in various ways. In order to support data-driven research approaches within the scope of the Digital Earth project, we designed and implemented a conceptual, as well as a technical software framework. The core idea of the framework is to provide the components, modules and interfaces that allows for the transformation of a scientific workflow into its digital counter-part. The main concepts behind this are a distributed modularization approach and an overall structured and component-based development. This maximizes the re-usability of implemented methods, components and modules and enables the implementation of even complex workflow tasks and operations. Because of the strong focus on the transformation of scientific workflows we call our framework a ‘Scientific Workflow Engine’.

The key component linking all independent modules is a central message broker instance. The described modularization allows the realization of almost all possible scenarios. For example the usage of an arbitrary backend module (e.g. providing a specific machine learning method) by a multitude of frontends as shown in figure 1. Our approach also allows the chaining of compatible modules as depicted in figure 2. Since each module is a more or less independent piece of software, we support a wide variety of commonly used programming languages.

 

Fig1: Single back-end module used by multiple front-end applications

 

Fig2: Processing chain of multiple back-end modules using different programming languages

 

We used this engine to create the Flood Event Explorer prototype, that combines multiple scientific workflows in order to address a common overarching research question in the domain of flood events. Moreover, we realized multiple generic stand-alone modules, implementing specific scientific methods, that are applicable to various problem sets. Further details on the implemented workflows and their individual methods can be found on the respective workflow description page.

 

CONTACT

Daniel Eggert Section Remote Sensing and Geoinformatics, GFZ, email: daniel.eggert(at)gfz-potsdam.de