Output details

Refinement of artificial intelligence-based systems for diagnosing and predicting river health

<24> This demonstrates the merits of our AI techniques, embedded into operational systems, and their underlying models. The work provided the Environment Agency (EA) with diagnostic and scenario testing support for identifying and addressing pressures on rivers, and they underpinned its work towards meeting the EU’s Water Framework Directive (WFD) to achieve at least ‘Good’ quality for UK rivers, by 2015. The extensive dataset constructed in the project has been used by others, such as for WFD inter-calibration, and assessment of ecological and chemical impacts. This work underpins the ‘Computer-Based Methods for Diagnosing and Predicting River Health’ impact case study.

With EA funding (£275k, 2004-2009), this 255-page research report evaluates a large project with three main axes:

• Model development, validation and software implementation, to inform the development of regulatory standards for improving ecological quality [Report Section 2]

• Cleaning and preparation of data to produce a very large research dataset [Sections 3 to 7].

• Enhancement of AI models through model development, validation and software implementation [Sections 8 to 12].

Besides contributing to the research concepts used above, Trigg also completed the day to day work for the three parts above, including software development in all aspects and the necessary AI development.