Environment Agency: Interpreting Complex Data to Help Clean up Rivers

It's not unusual for a business or organisation to quite literally have more information than they can handle.  When that happens, don't despair, as this example shows.

The Environment Agency is tasked with monitoring and improving the quality of all river water in the UK by 2015. Rather than simply testing the water, the only way to understand exactly what is going on in a river is to analyse the biological community that lives within it. The job of interpreting this data is a complex task that usually falls to an expert ecologist. Experts, however, are a rare commodity and their findings often subjective or inconsistent.  And the water quality can often vary wildly in between sampling, which makes assessing data less than straightforward.

With the 2015 deadline steadily approaching, and a great deal of data being generated from rivers all across the country, the Agency needed a more consistent, objective and available method of assessing the true water quality represented, and any issues, without incurring vast costs.

The response of the team at our Centre of Intelligent Environmental Systems was to produce two expert systems capable of interpreting river biological monitoring data. The two systems were designed to model two different methods of approaches to interpreting biological data that an expert ecologist uses, namely memory and reasoning.

‘Pattern matching’ replicated the use of memory when recognising previously seen patterns, helping cluster and make sense of large, confusing data-sets.  The second technique, a ‘plausible reasoning system’, was used to effectively replace an on-site expert, assisting Agency staff in their work of interpretation.

Between them, the two new systems have greatly assisted the Environment Agency in working towards its forthcoming deadline.

From the viewpoint of other businesses, both these techniques are entirely applicable to commercial databases where an organisation is having trouble understanding what to do with a particular opportunity - when data they hold may not have been collected for that exact purpose.  If there is a parallel here with any aspect of your own work, you can explore the help available at Staffordshire University via a conversation with us.

Our vast knowledge base and innovative thinking has helped businesses and organisations such as this to achieve greater success. Start finding out how we might help your business

• Downloadable version of this case study (PDF, file size: 258.17KB)