Feeds

Climate change: looking for a haystack, not a needle

Micro versus macro

Next gen security for virtualised datacentres

Column Weather forecasting. It's the archetypal "science doesn't know jack!" cop-out for mystics who think climatology isn't a science.

Amuse yourself. Google for "climate change" and "weather forecasts" together. Among the many rants with LOTS OF CAPITALS and many exclamation marks, you'll find over 1,000 comments to Heidi Cullen's blog on the Weather channel. Here's a typical sample:

I have to wonder why you, Dr Cullen, would expect the public to believe you and others at the weather channel about your 'forecast' of the earth for the next 100 years when you are only correct 30% of the time about the forecast of the weather for the next week? Please just stick to providing us your incorrect forecasts of the weather and leave the politics to the morons in the DC.

This pops up so often, I have to ask why? Why do so many people try to make a connection between weather forecasts and climate forecasts? "Because," you'll be told, "at the end of the day climate is made up of weather? A case of micro and macro."

Well, yes. And so, why don't we play a little game, where we think of things where the micro is hard to predict?

A good example: take a gram of radioactive material. It's a hitherto unknown element, and it has a radioactive half life of one hour.

Physics experiments have shown that in any radioactive material you cannot predict the moment of decay of any particular atom of that material. And yet, you can be absolutely sure that after an hour half the atoms in a lump of material with a one-hour half life will have decayed.

Let's think of something else. How about the monsoons of India? You can never say which day the monsoon will start, and there are even years when it fails. And yet, you can be absolutely sure that in a decade there will be at least nine monsoons in India.

Take climates of opinion. I can't tell you when the next paedophile scare will occur, but I can be sure that over the next 12 months there will be several. I can predict the same about "mobile phones are rotting our brains" scares.

Right now, it's summer. I can't tell you which day will be the hottest of the year, but I can be sure that the hottest day this summer will be about ten degrees above the hottest day from December through February.

Knock over an ant hill. You'll see ants run everywhere; so now, bring your foot down in their midst. I can guarantee you'll kill many on the macro level; but neither you nor I have any way of looking at the swarm in advance and saying which, at the micro level, will live or die.

Start a business. Will you succeed or will the fledgling enterprise go titsup in its first year? Millions are thrown away each year by investors who would love to be able to say "yes" or "no" in advance, and yet year on year roughly 11 out of 12 startups die before they reach their birthday - as predictable as sunshine in the Sahara.

If the economic climate changes, of course, and recession sets in, you can see that change reflected in the increasing number of business failures. That's predictable, but does the knowledge that it will happen help you predict which businesses will survive?

Twenty horses line up at the starting gate. We can predict with astonishing accuracy that at least 19 of them will be past the finish line half a mile away within a minute or two of the "off", but anyone who pretends to know which of the horses will be first across the line is out to get your cash.

A storm hits London. Winds shake the windows and doors. Any actuary will tell you that the following day he'll get claims for roof damage. Can he tell you which house will be hit by the gust?

It's become something of a challenge. We can think of any number of examples where, with little skill, the macro situation is predictable even by the feeblest intellect, but where the micro events which make it up are going to elude the cleverest forecaster.

Even people who deny that climate is changing work on the same basis: they look at average highs and lows over an extended period and try to see what the macro trend is. Then, despite their claim that it can't be predicted, they say: "There isn't any significant change," it's not happening.

Individual years will vary on the micro level. It's impossible to say "next year will be an El Nino", or "in two years, a volcano will fill the Northern hemisphere with dust, causing unusually low temperatures", but average your observations out over a sufficiently macro level, and you'll have a good picture of what is going on.

And there's a final point worth remembering. Although we mock them when the rain comes a few hours late, or when clouds don't drop a shower on our particular corner of the country, the truth is that weather forecasts are much better than people acknowledge. What we can't do is predict long-term weather. But every day in every city, millions get up and look at the sky, and say: "Hm, it's bright and sunny. I don't need an umbrella to get to the station," and most of the time they're quite right. What will happen is probably that whatever is happening now will carry on happening.

Even in April shower periods, it works rather like that. You can be pretty sure that the weather will remain "intermittent short showers" and sensible folks predict it and take their wet weather gear.

Climate predictions work on a longer scale. We may not be able to say "the next Ice Age will begin in 2,000 years" with any hope of being right, but when someone looks at the figures and says "the trend shows the normal unpredictable short term variation, but overall, temperatures average one degree more than they did" then the sensible forecaster will bet on the trend continuing in the short term. And in climate, "the short term" means more like "the next century or two" - over which period, it's probably pretty predictable within a range.

So, what instinct compels quite smart minds to reach for this comfort - "weather forecasters get it wrong!" - when it comes to climatic predictions?

And, if you want to amuse yourself, can you think of examples where it works the other way? - where averaging out the many short-term variations in something doesn't give a good picture of the longer term trends for the mass?

Drop us a line via the "contact the author button", or post your comment for all the world to see. ®

Secure remote control for conventional and virtual desktops

More from The Register

next story
Our LOHAN spaceplane ballocket Kickstarter climbs through £8000
Through 25 per cent but more is needed: Get your UNIQUE rewards!
LOHAN tunes into ultra long range radio
And verily, Vultures shall speak status unto distant receivers
EOS, Lockheed to track space junk from Oz
WA facility gets laser-eyes out of the fog
Volcanic eruption in Iceland triggers CODE RED aviation warning
Lava-spitting Bárðarbunga prompts action from Met Office
NASA to reformat Opportunity rover's memory from 125 million miles away
Interplanetary admins will back up data and get to work
LOHAN Kickstarter breaks NINETEEN THOUSAND of your EARTH POUNDS
That's right, OVER 9,000 beer tokens - and counting
Major cyber attack hits Norwegian oil industry
Statoil, the gas giant behind the Scandie social miracle, targeted
prev story

Whitepapers

Endpoint data privacy in the cloud is easier than you think
Innovations in encryption and storage resolve issues of data privacy and key requirements for companies to look for in a solution.
Implementing global e-invoicing with guaranteed legal certainty
Explaining the role local tax compliance plays in successful supply chain management and e-business and how leading global brands are addressing this.
Advanced data protection for your virtualized environments
Find a natural fit for optimizing protection for the often resource-constrained data protection process found in virtual environments.
Boost IT visibility and business value
How building a great service catalog relieves pressure points and demonstrates the value of IT service management.
Next gen security for virtualised datacentres
Legacy security solutions are inefficient due to the architectural differences between physical and virtual environments.