A large part of what the public sees concerning climate change involves graphs. Graphs can be an excellent visual aid. Graphs can also push people in the direction of the conclusions one wants them to reach.
This is probably my favorite x-axis label:
Since climate change is not actually based on global mean temperature but rather anomalies, labeling the axis “Arbitrary Baseline” is appropriate.
This is from Skeptical Science, to illustrate the “proper” way of looking at temperature anomaly data. However, the skeptic’s use of a stepped progression is a legitimate statistical technique and is appropriate in some cases. The bright red of the realist’s line also draws one’s eye to that line, while the blue against the green data lines and points serves to obscure the skeptic lines. There is a very real psychology to presenting data in graphs.
The scaling on a graph can make a very strong, visual statement. Notice how much more dramatic the first graph looks. If we elongate the y-axis and shorten the x-axis, the increase in temperature becomes much less noticeable. The data is exactly the same, but the first graph is more likely to convince the general public that climate change is real. The choice of the graph scales probably depends on what point the author is trying to make.
A couple of graphs showing warming and cooling over the ages:
These graphs reinforce the idea that climate has always changed. Both graphs are unlikely to convince people current warming is significant or out of the ordinary. The striking color contrasts make the cool/warm periods really stand out. The graph scales also add to the feeling of widely varying temperature swings.
Ignoring the line through the data, and looking at only the data, predict what comes next:
A appears to be headed toward cooling.
B shows a definite downward trend.
C show a dramatic increase.
In reality, the data on Graph A returns to an upward trend after the drop.
Graph B starts back up, then down then up again.
Graph C is the peak of a warming period, after which the data drops down again.
I include this to illustrate that even with line of best fit or whatever method one uses to smooth the data and then graph it, the graph can certainly lead to seeing a trend that goes in a wrong direction. It’s very easy to predict data that won’t happen for years or decades. However, it’s not so simple when the data is already there and you are predicting from past known data to current known data.
This is the actual full-length graph:
Then there are the graphs that represent what appears to be the same period and is reporting the anomalies in the temperature, but the graphs look totally different. Again, what the author wants to highlight seems to be the reason for the choice of graph styles. This also is in part why the average person on the street thinks climate change is simply fabricated. I am not saying it is, nor that both graphs don’t have reasons for the choices of graph styles. It’s just not evident without much deeper digging why these two graphs look completely different and appear to be predicting different outcome.
Then there are graphs like this one:
Much like the Marcott graph used in the example above this one (graph above right), this graph looks fabricated. We are looking at 450,000 years of information based on proxy data and the correlation is nearly perfect. It is unlikely that such perfection could actually occur with such a long period of time involved and proxies for all variables. The first reaction to this type of graph is to ask for the data and the methods used to find out why nature is suddenly so perfectly coordinated in this case. Very few variables in nature that are as complex as climate correlate perfectly.
This is the last graph. I include it as an example of a useless presentation of far too many variables at once. While the graph may actually represent useful data, one is hard-pressed to see what any of that data actually is. I am uncertain that the graph was actually intended to convey information. However, if it was, it failed.
Since we are dealing with proxy data in much in much of climate change science and some of these graphs, I thought I would ask:
Climate change advocates expect people to accept that changes as small as .1 degree change in the global mean temperature can be calculated using tree rings, ice cores, etc. This is necessary because there is no instrumental data (or the data is said to be corrupt or problematic) for the periods in question.
Climate change questioners have been told they cannot understand the complexity of the science, ie there is no direct data for the theory. They have been asked to accept the word of experts. When questioners instead look at “proxy” data—the behaviour of the scientists, the politicalization of the subject, the predictions that failed, the attacks on the questioners and name-calling of these persons—advocates scream “foul”.
One question that often comes up is why proxy data is not used all the way through the article and analysis. Why do we not use proxies up through 2013?
Are there studies that show proxies now match temperature readings now? Generally, the answer seems to be “no”. I have located one possibility but it is 200 pages long and will take some time to wade through.
Questioners proxies are based on scientific methods and procedures. Science is about data and results, not activism, shouting people down and declaring one’s self supreme authorities. Advocates say this is not the case and that experts are who we should follow.
Until such time as either groups proxies are proven incorrect, the proxies will be used, I am sure.