Selected Presentations
Home > Presentations > Abstract details
Title
ENVIMODEL: SCIENTIFIC WORKFLOWS AND WPS GEOPROCESSING FOR CLIMATE CHANGE
Abstract
We present EnviModel, an open source web interface aimed at supporting interdisciplinary research for climate change studies. EnviModel provides a platform for developing and sharing environmental models from spatial temporal data through scientific workflows and WPS geoprocessing. The platform is based on open source solutions and uses OGC standards for geodata, metadata and web services. Merging a WebGIS interface and workflow solution originally developed for bioinformatics application, we aim at supporting reproducible science in geoinformatics. Complex pipelines (e.g. risk models depending on climate change scenarios) can be built chaining together geo-computing tools, saving intermediate and final results. Analyses and pipelines can be shared, repeated or collaboratively modified by users without strong computational skills, with a potential to support transformative research. EnviModel is being developed as a module of the Enviro tool, the computational platform of the ENVIROCHANGE project to implement and share environmental risk models on the potential impact of climate change on agriculture. The module is based on ExtJS, WPS server technologies (52North, pyWPS) and Galaxy, a web based editor for scientific workflows (http://g2.bx.psu.edu). Models are translated into WPS by using Geotools, Sextante and GRASS geoprocessing functions. The Galaxy platform, is written prevalently in Python, has been extended to execute requests, orchestrate WPS and WPS chaining. The system can be further extended using Python scripts while inputs and outputs are defined using an external XML file. On-the-fly calculation of risk scenarios is a main EnviModel's feature. A meta information environment for the description of models and scenarios completes the interface, with the possibility of adding technical documents, scientific papers and detailing inputs, outputs of the chain. The grape phenology model, the Huglin index and the Botrytis cinerea risk index are already implemented and they will be presented as examples. ENVIROCHANGE is funded by the Autonomous Province of Trento, Italy.
Authors
Riccardo De Filippi - Fondazione Bruno Kessler
Davide Albanese - Fondazione Bruno Kessler
Shamar Droghetti - Fondazione Bruno Kessler
Marco Grimaldi - Fondazione Bruno Kessler
Vanni Tomasi - Fondazione Bruno Kessler
Calogero Zarbo - Fondazione Bruno Kessler
Ilaria Pertot - Fondazione Edmund Mach
Cesare Furlanello - Fondazione Bruno Kessler
