Met Forecast Of Summer Downpours Ignores Evidence

By Paul Homewood 

 

http://www.bbc.co.uk/news/science-environment-27624478

 

The BBC report on the latest study by the Met Office & Newcastle University:

 

Global warming will lead to a significant increase in extreme summer downpours in the UK, a study suggests.

The Met Office and Newcastle University researchers say there could be five times the number of "extreme rainfall events" exceeding 28mm per hour, under extreme warming projections.

This would cause "really severe" flash flooding in many parts of the UK, according to the scientists.

However, they caution that this result is based on only one computer model.

Flash flooding in Britain has had devastating impacts on communities in recent years.

In Boscastle, north Cornwall, about 200mm (8in) of rain fell in four hours in August 2004 causing a 3m wall of water to sweep through the village.

In the summer of 2012, in Newcastle, the equivalent of a month’s rain fell in just two hours, causing widespread flooding in the city.

Super models

Researchers have struggled to work out how global warming might affect these types of events.

Until now, their climate models have not been good enough to work out the effect on extreme hourly rainfall in the warmer months.

To improve the resolution of their model, researchers in this latest experiment used 1.5km grid spacings instead of the normal 12km.

To gain this extra clarity, the Met Office supercomputer was employed for nine months to run the simulations. Even then, they could only model the southern half of the UK.

"Most people would be familiar with this model," Dr Lizzie Kendon, the report’s lead author told BBC News.

"It is the same one that is used for the weather forecasts on the BBC, so it is incredibly realistic and it represents these very intense convective-type storms that haven’t been captured before."

The researchers used both the low resolution and the high resolution models to examine the climate patterns that have occurred in recent years and to look ahead to what might happen at the end of this century.

They assessed the period up to the year 2100 using the most high-end climate projections from the Intergovernmental Panel on Climate Change.

When they looked at rainfall patterns for the winter months, they found that both the 12km and 1.5km grid models showed an increase in rainfall.

Both models found that summers in the future would be drier overall.

However, when it came to intense downpours, defined as more than 28mm per hour, the higher resolution model saw a significant increase.

 

The study suggests that flash flooding events in the summer could become much more common in the UK

It found that there could be up to five times the number of events per hour than we see currently.

"It is dry periods interspersed with these very intense downpours, and we are talking about thresholds of 30mm and above in an hour over quite a large area here, which would be associated with really severe flash flooding," said Dr Kendon.

The researchers stress that this is the result of just one model run and it is not a definitive forecast.

Temperatures may not rise at the level used in the model.

However, the scientists believe that their work shows that global warming will make downpours a more frequent event in British summers.

"From this model experiment and consistent with our theoretical understanding, we have quite a bit of confidence in this result."

Prof Hayley Fowler, from Newcastle University, who is another author of the paper, said the new study was an important step to understanding the flooding risks of the future.

She hopes that other research groups will try to replicate the study.

"The next steps are to see if these changes are consistent with observed trends in summer rainfall extremes and changes projected by climate models in other parts of the world," she added.

 

 

 

Despite the warnings in the study that this just the result of one, solitary model, and that the most high-end climate projections have been used, it is no surprise that the BBC use the alarmist headline “Global warming will lead to a significant increase in extreme summer downpours in the UK”.

 

 

I also note they give the example of Boscastle, which suffered from just such a flash flood in 2004. What they fail to mention is that a very similar flash flood, every bit as bad as Boscastle’s, occurred a few miles along the coast at Lynmouth in 1952 after 9” of rain fell. This tells us two things:

1) The Boscastle flood was not unprecedented.

2) Such floods are still very rare.

 

It is also probably worth pointing out that summer temperatures in England are hardly increasing in leaps and bounds!

 

 

http://www.metoffice.gov.uk/climate/uk/summaries/actualmonthly

 

 

As stated, the study is based purely on one model, so how does it compare with recent trends and observations?

There is virtually no hourly rainfall data available, to use for long term comparisons. However, daily data is available, which should give us a good idea whether summer downpours are getting worse. It is, after all, reasonable to assume that, if “extreme rainfall events exceeding 28mm/hour” are becoming more common, then we should also expect to see a like increase in “extreme rainfall days”.

We can test this by having a look at a few sites, starting with Oxford, pretty representative of the area the study is looking at. Figure 1 plots the days with 30mm and over for summer months, available from Met Office data since 1930.

 

Figure 1

 

There is no evidence of any trend to heavy rainfall days becoming either more common or severe. The record for the wettest day was set in 1968, and extreme rainfall days were most common between 1950 and 1980, as Figure 2 illustrates.

 

Figure 2

 

We can repeat the exercise for Shawbury, in the West Midlands, and Heathrow.

 

Figure 3

 

Figure 4

 

 

Both show the same pattern, with the most extreme rainfall occurring between 1950 and 1980, suggesting that there is a pattern here. This is significant, because for most of that period, with the exception of the heatwaves of 1975/6, these decades experienced largely colder than normal summers.

There is therefore an implication here that extreme rainfall in summer is associated with cold, and not hot, conditions.

There is one more test we can do. Figure 5 shows the average rainfall/rainday, for the summer months in England South.

The Met Office only started keeping rainday data in 1961, so we have no figures prior to then. Nevertheless, the 1960’s and early 70’s show up as giving the wettest days. The purple trend line gives no evidence of any increase.

 

Figure 5

http://www.metoffice.gov.uk/climate/uk/summaries/datasets

 

 

 

Of course, we cannot automatically assume that hourly rainfall behaves in exactly the same fashion as daily rainfall. However, the data we do have suggests that when it did rain, in the summers of the 1950’s and 60’s, it tended to be heavier.

 

Two other things are worth mentioning.

 

1) The Met Office keep a page detailing record rainfall, etc.

The most recent record for any short duration was set 1989.

 

http://www.metoffice.gov.uk/public/weather/climate-extremes/#?tab=climateExtremes

 

2) In their Press Release, the Met Office mention winter rainfall:

In winter it is the daily or multi-day rainfall totals that are important, because we tend to get steady, long-lasting periods of rain from large scale weather systems – similar to those seen during the winter floods of 2013/14.

 

At the time, Julia Slingo was talking about intense rainfall, but this statement rather confirms what I have previously pointed out, that rainfall totals last winter were so high because of sheer persistence, and not intensity.

 

Conclusions

When forecasting the future effects of a warming climate, surely the first thing to look at is what effects a warmer climate has had in the past. Daily data will not necessarily provide the same results as hourly, but, nevertheless, that is the best information the Met Office have.

My analysis only covers three stations; I would like to use more, but the Met Office now operate a “fair usage policy”, which restricts the release of such data to small chunks. (The data I have used was sent to me last year, before this policy was put into effect). There is no reason, however, why a fuller coverage of sites could not have been used by the authors of this study.

While climate models have their uses, is it too much to expect researchers to spend a bit of time actually looking at the data as well?

 

 

Footnote

It was only back in March that another Met Office paper, “Drivers and impacts of seasonal weather in the UK”, stated on page 22,

“There is no clear signal for how the wettest days in summer may change”

 

Sources

All the daily rainfall data used were supplied to me by the Met Office.