Milestones

State of art: actual results

Results of the first stage

Within the first stage of the project different approaches for Grid system development were analyzed. The object-oriented simulation model of distributed information system for satellite data processing was developed. Based on this model basic infrastructure of Grid system for satellite data processing was deployed using resources of Space Research Institute of NASU-NSAU and Institute of Cybernetics of NASU. Component interaction was elaborated and tested using as testbed the problem of modeling of dynamic processes in atmosphere using WRF model. The service for forecasting meteorological parameters using WRF model was run in operational mode. The results of the visualization of developed approaches are accessible via Internet by dos.ikd.kiev.ua.

During the first stage the methods for applied problems solving were also developed. In following stages these methods will be implemented in the Grid system in the form of services. In particular, experimental research of neural network methods for interplanetary shock prediction based on data from ACE satellite was carried out. Variation problems and weak-statement problems were built up for description of stable vertical and profile fluid filtration in multi-component environment. Differential methods for interpolation of meteorological images based on optical flow were improved. A new dynamical-information approach to the forecasting of solar coronal ejection mass has been proposed.

Results of the second stage

Within the second stage of the project the basic GRID infrastructure for satellite data processing was developed. Grid infrastructure integrates information and computational resources of Space Research Institute and Institute of Cybernetics. The infrastructure also includes the newly installed EUMЕТCаst broadcast system for operational reception of environmental data from satellites.

Based on the system analysis a structural and functional decomposition of the system has been made. Using object-oriented approach the tasks that are solved in the Grid system have been formalised. Complex jobs have been decomposed on elementary tasks. Sequence diagram have been developed for complex jobs using UML. Developed models were implemented in simulation package GridSim. Initial class diagram of GridSim have been extended. Job workflows were also implemented in GridSim using developed models.

During the second stage the algorithms and services for applied problems solving were also developed. In particular, a parallel algorithm for meteorological parameters forecasting using mesoscale model WRF and nested nets was developed. New algorithm for the parameter identification of nonlinear discrete dynamic system (NARMAX model), which describes the connection between time series and Dst index of space weather was developed. The numerical algorithm for two CME models as isolated plasmoid was created and a series of calculations was carried out. It was shown that the developed models predict CME cloud traveling time to the Earth more precisely, than it was known.

A service for onset detection of inter-planetary shock in solar winds and forecasting of arrival time of inter-planetary shocks on the Earth using data from ACE satellite was developed. The algorithm for numerical analysis of fluid motion in multi-component soil environment was developed.

Results of the third stage

On the final third stage main works were directed to the overall system testing: infrastructure and services. Special attention was paid to the promotion of Grid system use among end-users and scientific community. All proposed methods were implemented in the form of services in the Grid infrastructure. The obtained results of system testing and performance showed its applicability for the solution of complex problems of environmental monitoring.

In order to monitor and submit jobs to the Grid system a Grid portal was developed and accessible via Grid Portal. Much of the work was done to integrate developed methods with third-party systems. In particular, developed services for space weather monitoring are visualized on the portal of Ukrainian Space Weather System (http://www.spaceweather.org.ua).

Services for meteorological parameters forecasting were among first 15 services that were registered to GEOSS Architecture Pilot Infrastructure as contribution to GEOSS from Ukraine. Results of the projects were published in scientific journals and discussed on numerous national and international conferences and workshops. Based on the results obtained during this project participating institutions were invited to take part in activities carried out by International Charter "Space and Major Disasters". The results obtained within the project could be used as a basis for development of operational services and further commercialization of the resultss.