Statistical explainer: average temperature increases can be deceiving

12 05 2015

Beating-the-Heat-Without-PowerOver the years I’ve used a simple graphic from the IPCC 2007 Report to explain to people without a strong background in statistics just why average temperature increases can be deceiving. If you’re not well-versed in probability theory (i.e., most people), it’s perhaps understandable why so few of us appear to be up-in-arms about climate change. I posit that if people had a better appreciation of mathematics, there would be far less inertia in dealing with the problem.

Instead of using the same image, I’ve done up a few basic graphs that explain the concept of why average increases in temperature can be deceiving; in other words, I explain why focussing on the ‘average’ projected increases will not enable you to appreciate the most dangerous aspects of a disrupted climate – the frequency of extreme events. Please forgive me if you find this little explainer too basic – if you have a modicum of probability theory tucked away in your educational past, then this will be of little insight. However, you may wish to use these graphs to explain the problem to others who are less up-to-speed than you.

Let’s take, for example, all the maximum daily temperature data from a single location compiled over the last 100 years. We’ll assume for the moment that there has been no upward trend in the data over this time. If you plot the frequency of these temperatures in, say, 2-degree bins over those 100 years, you might get something like this:


This is simply an illustration, but here the long-term annual average temperature is 25 degrees Celsius, and the standard deviation is 5 degrees. In other words, over those 100 years, the average daily maximum temperature is 25 degrees, but there were a few days when the maximum was < 10 degrees, and a few others where it was > 40 degrees. This could represent a lot of different places in the world.

We can now fit what’s known as a ‘probability density function’ to this histogram to obtain a curve of expected probability of any temperature within that range:


If you’ve got some background in statistics, then you’ll know that this is simply a normal (Gaussian) distribution. With this density function, we can now calculate the probability of any particular day’s maximum temperature being above or below any particular threshold we choose. In the case of the mean (25 degrees), we know that exactly half (p = 0.50) of the days will have a maximum temperature below it, and exactly half above it. In other words, this is simply the area under the density function itself (the total area under the entire curve = 1). Read the rest of this entry »

Scaring our children with the future

21 01 2013

frightened childI’ve written before about how we should all be substantially more concerned about the future than what we as a society appear to be. Climate disruption is society’s enemy number one, especially considering that:

  1. all this unprecedented warming is happening on a template of highly degraded land- and seascapes. Extinction synergies (more extinctions than would otherwise be predicted by the simple sum of the different pressures) mean that climate change exacerbates the extinctions to which we are already committed;
  2. we show no sign of slowing emissions rates, partly because of the world’s ridiculous refusal to embrace the only known energy technology that can safely meet emissions-reduction requirements: nuclear power;
  3. there are 7 billion hungry, greedy humans on planet Earth, and that number is growing;
  4. scientific evidence denial, plutocracy and theocracy are all on the rise, meaning that logical, evidence-based decision making is being progressively tossed out the window.

That’s probably the most succinct way that I know of describing the mess we are in, which is why I tend to be more of a pragmatic pessimist when it comes to the future. I’ve discussed before how this outlook makes getting on with my job even more important – if I can’t reduce the rate of destruction and give my family a slightly better future in spite of this reality, at least I will damn well die trying. Read the rest of this entry »

Tropical forests cooking their biodiversity

5 05 2011

Another ‘hot’ essay by Bill Laurance recently published online by Yale Environment 360 (a publication of the Yale University School of Forestry & Environmental Studies). Bill asked me to relay it on, so here it is in full:

Much attention has been paid to how global warming is affecting the world’s polar regions and glaciers. But a leading authority on tropical forests [that would be Bill] warns that rising temperatures could have an equally profound impact on rainforests and are already taking a toll on some tropical species.

On Jan. 12, 2002, in the Australian state of New South Wales, biologist Justin Welbergen was observing a colony of flying foxes for his Ph.D. research. The temperatures that day on Australia’s subtropical, eastern coast reached record highs, soaring to 42.9 ° C (109 ° F) at the weather station closest to Welbergen’s study site — nearly 8 ° C higher than the average summer maximum temperature.

The flying foxes, or giant fruit bats, normally just doze in the treetops through the day, but on this afternoon they were fanning themselves, panting frantically, jostling for shady spots, and licking their wrists in a desperate effort to cool down. Suddenly, when the thermometer hit 42 ° C, the bats began falling from the trees. Most quickly died. Welbergen and his colleagues counted 1,453 flying foxes that died from the heat in one colony alone. The scorching heat that day killed at least 2,200 additional flying foxes in eight other colonies along a 250-kilometre stretch of coastline. All the deaths occurred in colonies where temperatures soared above 41.7 ° C. Read the rest of this entry »


Get every new post delivered to your Inbox.

Join 7,813 other followers

%d bloggers like this: