NEW ways to measure the amount of water rushing down rivers and streams that could help prevent flooding in future have been developed by Scottish scientists.

A team of researchers from St Andrews University examined the streamflow and channel capacity of rivers in the US to monitor trends in the number of times they burst their banks.

The researchers say the development, which quantifies flood frequency, can contribute to flood management systems and help estimate insurance costs more accurately.

With flood risks set to grow as climate change accelerates, the St Andrews earth scientists say that their work - published online today by Geophysical Research Letters - is especially relevant to the UK.

The research was led by Dr Louise Slater during her PhD studies in the Department of Earth and Environmental Sciences at St Andrews University.

Dr Slater said: "In assessing and analysing flood risk, channel capacity in streams and rivers has generally been assumed to remain relatively constant over set periods of time.

"We now know that trends in flood hazard are not just about hydrology. Compared to previous studies, we took a new approach and examined 401 rivers throughout the USA to compare the contributions of channel capacity and flow frequency to historical flood hazard and analyse their interactions.

"Our findings are especially relevant to the UK, given the recent controversy surrounding the causes of flooding in the Somerset Levels."

Increases in flood damage in recent decades are a global problem, impacting the environment and resulting in significant economic losses.

Dr Slater was supported in her research by co-authors Dr Michael Singer of the Department of Earth & Environmental Sciences at St Andrews University, and Professor James Kirchner of the Department of Environmental Sciences, ETH Zurich.

Dr Singer continued: "Flooding is a major hazard to lives, infrastructure and economic prosperity, but trends in flood hazard are currently poorly understood. By developing new methods to assess flood frequency, we have demonstrated that accurately quantifying changes requires the separate accounting of both streamflow and channel capacity."