The Krebs manoeuvre

Lord Krebs, the chairman of the adaptation committee of Lord Deben's Committee on Climate Change, has issued a report on how we are doing on preparing for the perils of climate change. The Independent has picked up on one of the themes of the report, noting breathlessly that British agriculture is in for a torrid time of it, with plants withering through lack of rain.

Within a decade, farmers could face a water shortfall of 115bn litres a year – almost half of the 240bn litres they currently use – with the south and east of the country, where most crops are grown, likely to be hit particularly hard. This could make it difficult – and more expensive – to grow water-intensive crops such as potatoes, carrots and fruit, the CCC warns.

Now future rainfall projections have been touched on from time to time at this blog, and once you get into the detail, it quickly becomes clear that nobody really has much of a clue whether it's going to get wetter or drier.

So how do we get to the Independent's predictions of apocalypse? If you go to the report itself you can indeed find the 115bn litre figure quoted. Here's the context:

There are larger uncertainties around future projections of rainfall in England than for temperature. Regional annual mean changes in rainfall could change by -10 to +10% by the 2020s (2010-2039), with changes in summer of between -30 to +20%, and changes in winter of between -10 to +30%. On balance, the evidence suggests that rainfall is more likely to increase in winter and decrease in summer, causing lower summer river flows.

This is about my understanding too. Overall, no change, summer could be lower or higher, perhaps on balance slightly lower.

However, when they develop their two scenarios for water supply, Lord K and his team come up with:

• a "less" water stress scenario with supply unchanged from today 120 Megalitres
• a "more" water stress scenario, with supply down to 100 Megalitres

It is the latter that produces a net deficit of 115bn litres once demand changes are taken into account.

This is rather odd: the range of summer rainfall changes coming out of the models is from a fall of 30% to a rise of 20%. Why do none of the scenarios consider such rises? But notice Lord K's use of what Mosher would call "high entropy language": not "high" and "low" water stress, but "less" and "more". See what I mean? Now read the small print:

• Scenario A (less water stressed) The supply scenario is based on the median of the wettest 1,000 sample runs from UKCP09 sampled data using the low emissions scenario (SRES B1)
• Scenario B (more water stressed) The supply scenario is based on the median of the driest 1,000 sample runs from UKCP09 sampled data using the high emissions scenario (SRES A1FI).

You will now see that the two water scenarios are based on different assumptions about greenhouse gas emissions. I think this means that the biggest increases in rainfall and the biggest decreases both take place for the high emissions scenarios - sometimes the model predicts a wetter future, sometimes a drier one. But by switching to the low-emissions scenario for the "less water stressed" scenario, Lord K and his team neatly avoid those horribly benign outcomes, leaving them free to base their advice on unremitting gloom and doom.

The predicted shortfalls in water are also dependent upon demand increasing. I don't propose to spend a lot of time on this side of the equation, but in essence it is predicted that the soil is going to be drier because of temperatures being higher.

There is high confidence that evapo-transpiration will increase rather than decrease because of the driving effect of higher temperatures. This is likely to lead to increased soil moisture deficits, particularly in the 2050s and beyond.

Which is odd, because just the other day, the former head of the Met Office, Lord Hunt, told us that lower summer temperatures were exactly what we should expect from global warming because it shifts the jet stream. Clearly Lord Krebs thinks that Lord Hunt is talking out of his hat.

In due course, the figures discussed above find their way to the Executive Summary (I assume the slight change in value is an error):

Our modelling suggests that, if current trends were allowed to continue, a gap could emerge between water supply and demand. In a dry year in the 2020s the gap could be nearly as large as total current agricultural abstraction of 120 billion litres per year.

Which I suppose is strictly true - their modelling does suggest that it could do this. But on the other hand, different figures are possible too - water supply is nearly as likely to increase and temperatures could even fall (if Lord Hunt is to be believed) in which case the situation would be completely different. But no matter, an argument has been put forward for a policy:

Reform of the abstraction regime must ensure that the price of water reflects its scarcity. This is required to incentivise improved irrigation efficiency and investment in on-farm storage, and contribute to ensuring sufficient water supplies in the future to meet growing agricultural demands.

It seems to me that rational policy development requires possible futures to be weighted according to their relative likelihoods. What the Committee on Climate Change have done is to ignore all benign outcomes and to base their advice only disaster scenarios. That is unacceptable and unforgiveable. Heads should roll.

Comments

[stone]

"Fresh-water supply shouldn't be a problem anywhere on the planet." I do not agree with Lynn Clark's point of view.