September 2016 - As part of the National Flood Resilience Review, the Met Office was asked to estimate the potential likelihood and severity of record-breaking rainfall over the UK for the next 10 years.
In order to provide this information, our teams used an innovative approach which combined our expertise in both weather and climate science. This work has contributed to the Review's report.
The flooding in 2013/14 and again in late 2015 was driven by large-scale frontal rainfall, a weather pattern that is often associated with river flooding and typically seen in the UK during winter months. The focus of our work was therefore on looking for synoptic weather patterns that give rise to large accumulations of rainfall that are likely to drive high river flows and flooding.
We also assessed whether climate change has played a clear role in recent rainfall and flooding events and concluded that natural variability is by far the dominant cause and will continue to be so for the next 10 years.
Consequently, we needed a way to estimate the wide range of possible weather and rainfall patterns which could affect the country. This needed to account for our highly variable climate, and include weather patterns which may not yet have been experienced in the observational record.
A map showing the standard Met Office climate regions across the UK. For the NFRR work, only regions in England and Wales were used. We used the latest version of the Met Office high-resolution climate model (HadGEM3). This new model has higher resolution than any used previously in climate prediction, with significant improvements in the way synoptic weather patterns, of relevance to UK flooding, are represented. This is the first time that the resolution of the climate model is regarded as 'fit for purpose' for a study of this nature.
Based on our archive of many possible realisations of the current climate, we were able to generate a large set of monthly rainfall events for 6 large regions covering England and Wales (shown in the map on the right). This event set, which consists of over 11,000 monthly rainfall scenarios, is many times larger than the set available from observations for the same period, 1980 to 2014 (i.e. 420 months).
This has allowed us to identify several hundred extreme monthly rainfall events that are greater than current existing rainfall records, but are regarded as plausible for the current climate - what might be termed 'black swan' events. We can then estimate the probability of certain levels of extreme rainfall occurring, which lie outside what has currently been experienced but are meteorologically reasonable.
Based on this model event set, the probabilities of exceeding certain rainfall thresholds, in excess of current observed rainfall records, were then calculated. The Met Office used the annual probability threshold of 1% to define a plausible worst case scenario (i.e. there is a 1% chance that for any month in a given winter, the rainfall over a large region will exceed the existing rainfall record by a certain percentage). While the model event set contains much more data than the observations, it's still not large enough to provide a robust estimate of risk for probabilities smaller than 1%.
The model event set suggests that there is a 1% probability in any winter of experiencing 15-35% more monthly rainfall over each of the large regions than has been seen in the observations to date. Another way of expressing this is to say that there is a 90% chance that such extreme rainfall won't be seen in one of those regions over the next 10 years. The event set also showed that there is around a 10% chance in any given year of existing monthly rainfall records, over any of the large regions, being matched and/or broken.
It is worth noting that these numbers are based on monthly rainfall for large regions of the country; it is entirely possible that on smaller space scales or for shorter time periods, the percentage exceedance of existing rainfall records may be higher (or lower). So we should not be surprised to see new rainfall records being made in the next few years.
With these estimates of how much additional rainfall might be plausible in the current climate, the Met Office worked with the Environment Agency to identify specific case studies with which to stress test their flood models and their extreme flood outlines. These were selected to capture differing hydrology across the country, and to focus primarily on recent extreme flooding events to ensure that any impact of climate change to date was factored into the tests. We also considered the combination of extreme river flooding with an extreme tidal storm surge.
Rainfall accumulations are very hard to measure because of their large spatial and temporal variability. Rain gauges only provide spot measurements and radar has difficulty in estimating rainfall in mountainous regions and when rain is very heavy. An alternative is to use the rainfall produced by the Met Office UK kilometre-scale weather forecast model, which we know performs well and contains the meteorological processes that drive extreme rainfall accumulations. This also allows us to provide rainfall data on fine time (every 15 minutes) and space (every 2 km) scales for each of the case studies, which we know is important for the hydrological response.
The rainfall data for each test case were then increased by a certain percentage in order to simulate a plausible extreme case. We combined the evidence from the climate model event set described above, with expert judgement drawn from our understanding of model limitations and a detailed analysis of past rainfall records, to define appropriate percentage uplifts for each case study. These ranged from 20% for North-West England to 30% for cases in Southern England. The percentage uplift in rainfall was applied uniformly in space and time to the whole rainfall scenario, and consequently represents a substantial increase in the volume of water entering the catchments. It can therefore be regarded as a robust test of the extreme flood outline.
These extreme rainfall data were passed to the Environment Agency, allowing them to feed these into their flood models and compute the likely flooding extent that would ensue. The results of these stress tests are presented in the NFRR report.
This work was reviewed by a scientific advisory group. This group comprised of experts from across the UK science community and was chaired by Sir Mark Walport, who is the Government Chief Scientific Advisor.
This new method, in which climate model simulations are used to explore the potential for meteorological 'black swans', is recognised as an innovative approach to estimating future environmental risks, such as flooding. These future environmental risks are associated with weather and climate variability, as well as climate change.
Last updated: 7 September 2016