Geophysical Research Abstracts
Vol. 12, EGU2010-14125-1, 2010
EGU General Assembly 2010
© Author(s) 2010
gProcess and ESIP Platforms for Satellite Imagery Processing over the
Grid
Victor Bacu (1), Dorian Gorgan (1), Denisa Rodila (1), Florin Pop (2), Gabriel Neagu (3), and Dana Petcu (4)
(1) Technical University of Cluj-Napoca, Computer Science Department, Cluj-Napoca, Romania
(dorian.gorgan@cs.utcluj.ro), (2) Politehnica University of Bucharest, Bucharest, Romania (florinpop@cs.pub.ro), (3)
National Reasearch Institute of Informatics, Bucharest, Romania (gneagu@ici.ro), (4) Western University of Timisoara,
Timisoara, Romania (petcu@info.uvt.ro)
The Environment oriented Satellite Data Processing Platform (ESIP) is developed through the SEE-GRID-SCI
(SEE-GRID eInfrastructure for regional eScience) co-funded by the European Commission through FP7 [1]. The
gProcess Platform [2] is a set of tools and services supporting the development and the execution over the Grid
of the workflow based processing, and particularly the satelite imagery processing. The ESIP [3], [4] is build on
top of the gProcess platform by adding a set of satellite image processing software modules and meteorological
algorithms.
The satellite images can reveal and supply important information on earth surface parameters, climate data,
pollution level, weather conditions that can be used in different research areas. Generally, the processing
algorithms of the satellite images can be decomposed in a set of modules that forms a graph representation of
the processing workflow. Two types of workflows can be defined in the gProcess platform: abstract workflow
(PDG – Process Description Graph), in which the user defines conceptually the algorithm, and instantiated
workflow (iPDG – instantiated PDG), which is the mapping of the PDG pattern on particular satellite image and
meteorological data [5]. The gProcess platform allows the definition of complex workflows by combining data
resources, operators, services and sub-graphs.
The gProcess platform is developed for the gLite middleware that is available in EGEE and SEE-GRID in-
frastructures [6]. gProcess exposes the specific functionality through web services [7]. The Editor Web Service
retrieves information on available resources that are used to develop complex workflows (available operators,
sub-graphs, services, supported resources, etc.). The Manager Web Service deals with resources management
(uploading new resources such as workflows, operators, services, data, etc.) and in addition retrieves information
on workflows. The Executor Web Service manages the execution of the instantiated workflows on the Grid
infrastructure. In addition, this web service monitors the execution and generates statistical data that are important
to evaluate performances and to optimize execution. The Viewer Web Service allows access to input and output
data.
To prove and to validate the utility of the gProcess and ESIP platforms there were developed the Green-
View and GreenLand applications. The GreenView related functionality includes the refinement of some
meteorological data such as temperature, and the calibration of the satellite images based on field measurements.
The GreenLand application performs the classification of the satellite images by using a set of vegetation indices.
The gProcess and ESIP platforms are used as well in GiSHEO project [8] to support the processing of Earth
Observation data over the Grid in eGLE (GiSHEO eLearning Environment).
Experiments of performance assessment were conducted and they have revealed that the workflow-based
execution could improve the execution time of a satellite image processing algorithm [9]. It is not a reliable
solution to execute all the workflow nodes on different machines. The execution of some nodes can be more time
consuming and they will be performed in a longer time than other nodes. The total execution time will be affected
because some nodes will slow down the execution. It is important to correctly balance the workflow nodes. Based