Distributed virtual experiments in water quality management.

Area of Science:

• Environmental Engineering
• Computational Science

Background:

• Increasing complexity of virtual experiments (VEs) and their models strains monolithic software performance.
• Computational bottlenecks limit the scope and efficiency of VEs in fields like probabilistic design and model calibration.

Purpose of the Study:

• To implement a framework for distributed execution of VEs to overcome computational limitations.
• To provide a stable, expandable, and platform-independent solution for complex VEs.

Main Methods:

• Developed the WEST distributed virtual experimentation (WDVE) framework using C++, XML, and SOAP.
• Designed WDVE for Master-Slave architecture enabling concurrent execution of sub-experiments on distributed nodes.
• Facilitated transparent transfer of experiment descriptions and required resources (models, input data).

Main Results:

• WDVE enables distributed execution of complex virtual experiments on heterogeneous computing resources.
• The framework is designed for stability, expandability, performance, platform-independence, and ease of use.
• WDVE offers a solution similar to Grid Computing but tailored for water quality management.

Conclusions:

• WDVE effectively tackles computational bottlenecks in complex virtual experiments.
• The framework enhances the feasibility and efficiency of VEs, particularly in water quality management.
• WDVE provides a robust and scalable solution for distributed scientific computing.