In my last post, I started to look at what comprises the modern phenomena of “climate science”. The definition I used had three main points:
1. It’s a science:
2. It’s about average conditions, not weather; and
3. It’s about average conditions over time.
Climate science is concerned with “average conditions”. We all know what average means. “He’s of average height.” “I had an average day.” etc.
The “average” or “mean” is a term from the branch of mathematics called statistics.
So climate science is about using statistics to analyse various climactic conditions. Which conditions?
Some of the obvious ones are:
· Wind speed and direction; and
· Rain, snow, hail and other precipitation.
Some other, less obvious ones include:
· Clouds of various types, shapes and configuration;
· Storms; and
· Total solar irradiance, the amount of the sun’s energy at various wavelengths, that hits the Earth.
The biggie, of course is temperature. We’re not told to be worried about polar bears getting wind burn, or getting fur-rot from too much rain. Temperature’s the problem, supposedly.
One of the things left out of the definition above is that all of these conditions can vary at different places on the planet’s surface. Obviously the temperature’s different at different places and, of course, at different times. There’ll be more about that in the next section.
While climate science deals with averages, it also deals with other statistical measures like minimums and maximums, as in “It was the hottest/ coldest/ wettest/ driest/ cloudiest/ clearest Tuesday afternoon in February since last month”. Maximums and minimums can also be treated with appropriate statistical tools.
Statements about averages, minimums, maximums and other statistical quantities can help us understand “the big picture”. They can also confuse and, in the worst cases, deceive.
The most important idea in statistics is this:
Everything we measure has an error. The temperature in my office wasn’t 29 degrees Celsius a moment ago. It was somewhere between a little less than 29 degrees and a little more than 29 degrees. Thermometers, and in fact all the stings we use to measure, have errors in their mechanisms that mean there’s always a bit of error or uncertainty in the measurement.
There are some statistical tools that allow us to estimate how much error is present in a measurement or set of measurements.
The third qualifier in the definition of “climate science” is that the average conditions are studied “over some time period”.
Sticking to the temperature example we might ask “What’s the temperature, where I’m standing, now?”, meaning over a very short period of time, say, a second.
Another, quite different question would be “What’s the average temperature of my office over the last year?”
I could actually answer this with some confidence if I got an electronic thermometer, connected it to my computer and got a reading of temperature every second for a whole year. That’s 365 x 24 x 60 X 60 = 31,536,000 seconds and, of course, the same large number of measurements.
To find the average I’d just add up the 31,536,000 measurements and divide by 31,536,000. (Of course, I’d get the computer to do the adding and dividing.)
I would certainly get a number. No doubt about that. How accurate would it be? That’s a topic for another post.
I could simplify this whole thing enormously by just measuring the temperature once or twice per day. Maybe I could measure it at midnight one day and noon the next or maybe whenever I got up or went to bed. I wonder if that would affect the uncertainty or accuracy of my measurement.
In summary, then, “climate science” is the study of the average values of conditions like temperature across varying periods of time.
In order to hypothesise an effect, like “Humans are causing global warming by burning fossil fuels.” We obviously need to make a before and after comparison.
Breaking the question down, we can ask:
“What was the average temperature of the Earth before wicked humans started burning fossil fuels?”
“What is the average temperature of the Earth now?”
“What’s it likely to be in one month?
“What’s it likely to be in one year?”
“What’s it likely to be in ten years?”
“What’s it likely to be in one hundred years?”
“What’s it likely to be in one thousand years?”
You get the idea.
Take note of the difficulties with all this:
· Thermometers have built-in errors that cannot be gotten rid of;
· We’re asked to measure the average temperature of the planet, but we only have thermometers in a relatively few places, mostly in North America. What do we do about all the places in between?
· How do we determine what the average temperature of the Earth was one hundred years ago? There were thermometers back there, but how accurate were they?
· How about one thousand years ago, before the invention of the thermometer?
It appears to me that “climate science” has set itself an impossible task.