An autumn snowstorm is chugging across the northeast US now. It brings to mind the anomalously cold events that have occurred in the US in the past couple of years, like the frosts in Florida. This, in turn, brings to mind the cacophony of recent news coverage questioning the reality of climate change. "Has Global Warming Stopped?" "If climate change is happening, why is the earth getting colder?" Is curbing greenhouse gas emissions "Treason"? I've often wondered, as I am wondering today: do changes in the weather effect the news coverage of climate change?
I am not a scholar of human nature. Still, questions like the one I've posed in the title seem important to climate science. Changes in climate and changes in weather are two very different processes that can be very weakly coupled in nature (depending on time scales). It takes many decades for the lumbering machinery of the climate to shift to another state. Weather, on the other hand, changes quite rapidly. I have lived in three different countries that boast the catch phrase, "If you don't like the weather, just wait five minutes!" Yet these two processes--weather change and climate change--can be very strongly coupled within the human mind. The confusion between weather and climate runs deep.
We often rely on news media for information on topics where we have no expertise. So I wonder: does the weather effect news coverage of the climate?
I'm not equipped to fully answer the question I've posed, but there is enough data freely available to at least begin to form an answer. I have put together a simple analysis to get at this question, using two time series. The first is the global mean surface air temperature change (provided by NASA). The second is an index of news coverage on the topic of climate change. This index is essentially the proportion of news articles in the US containing the term "climate change" (provided by the NewsLibrary.com database). Available news data begins in 1977, so that's where my time series start.
In figure (1), I have plotted the temperature change (top) and the climate change news index (middle). At a first glance, there might appear to be a strong correlation between the two (shown at the bottom). However, this strong correlation only says that there has been an increasing trend in both series. It says that temperatures have gone up over the past 33 years, and so has news coverage of climate change. Nothing surprising.
I'd like to know, however, if there is a correlation on a finer time scale. In other words, if there is a stand-out warm or cold year, is there any effect on the news coverage? One (of many) ways to test for this is to subtract the increasing trend out of the data and plot the residuals. When you do this, you get the flatter time series, shown in figure (2).
There is no significant correlation between these two time series. Yet if you look at them with a keen eye, you'll notice that they track each other fairly well until the year 2000, at which point they diverge. Breaking the time series up by decade tells an intriguing story. Figure (3) shows the correlations by decade--positive through 2000, then negative after 2000, and in all cases, statistically significant.
To put it into words, something changed in 2000. Before the turn of the millennium, warm years (i.e. warmer than the increasing trend) were associated with more news coverage of climate change. Cold years (again, cold relative to the increasing trend) were associated with less news coverage of climate change.
To me, this suggests that warm years were nagging reminders that this climate problem might be serious. Hence the news coverage. Have you ever been in a t-shirt on an eerily warm March day and heard someone remark sagely, "It must be global warming"? Warm spells are not unique to the past 20 years, and an odd weather event is not evidence of climate change. Yet they seem to remind many people of the climate change threat. (You would be wiser to listen to your grandfather, as he mumbles, "Why, when I was your age, we had to dig potatoes in summer, because the first snowfall usually came in late September, by cracky.")
What happened in 2000 is even more compelling. Suddenly it was cold years that drew the news coverage. Why would climate change make headlines in cold weather? I would hypothesize that this phenomenon is connected with the surge in climate change skepticism over the past decade. Cold years are fodder for this sentiment in the same way that warm were for the opposite view. I.e., "A snow storm in May? So much for climate change."
Mind you, this is only my hypothesis. We would need a deeper probing into the data to truly confirm or refute it. We would want to examine, for example, the context in which the phrase "climate change" appeared. Still, the evidence is compelling. It would be alarming enough to learn that our perceptions of climate change are driven by fluctuations in the weather, rather than sound climate research. It borders on disturbing that fluctuations in weather now seem to be driving a false perception.
On the other hand, it is possible that the increase in hurricanes, and the icy wind outside my door, are due to an angry Poseidon.
-Nick Record, signing off