Putting Graphs in Context

Too often data are presented in misleading ways. Certainly that has happened in the climate discussion, as alarmists try their very best to hype what they claim is a crisis. Data in context tell a different story.
New WUWT Global Temperature Feature: Anomaly vs. Real-World Temperature
Anthony Watts
One of the most frightening aspects of global warming, aka “climate change” is the graphs produced from temperature data for public consumption and trumpeted by an unquestioning and compliant media.…


Figure 5: NASA GISTEMP Data, plotted as a temperature anomaly, using a “normal” temperature of 57.2°F

Figure 6: NASA GISTEMP Data, plotted as absolute temperature, using a “normal” temperature of 57.2°F
Can you tell which graph visually represents a “climate crisis” and which one doesn’t?

Feel free to check my work – the Excel spreadsheet and the calculations are here:

GISSTEMP-in-absolute-masterDownload
To create the graphs above in Figures 5 and 6, I used the data from the Excel Sheet imported into the graphing program DPlot.

 

As you are about to demonstrate.

Neither of the graphs represent a "climate crisis", they both represent the steady increase in surface temperatures across the measured period. Whether that factual data is part of the evidence of or a factor in a "climate crisis" is a much bigger and more complex question. I can't be proven nor dismissed on the basis of any single graph (even if it is honestly and accurately presented).

The problem with your second graph is that it is presenting exactly the same data but that is being (intentionally) masked by extending the scale way beyond the scope of the data. The first graph is no more misleading than the second one, the key point is that the data they represent shouldn't lead to any definitive conclusion, they should just raise further questions and study; Could an increase on that scale have any negative consequences? Is the rate of increase peaking, steady or growing? Is this symptomatic of wider environmental factors? etc.

There is spin and misrepresentation from all extremes in this debate (as there are in most others). Simply producing more of the same is no solution to that problem.

 

Not sure what's extreme about a simple temperature record.

 

Bingo!

I remember seeing a video someone posted where the scale was in +0.2 C.

It was hilarious.

 

I never said it was. My point was that people on all extremes in debates misrepresent figures in the same kind of way and both of your versions of the graph could be used for that purpose.

 

I've pointing this out several times. Most warming fanatics lock onto the anomaly plot and go berserk. When you tell them the plot signifies a temp gain of about 1.2C a decade they call you a "denier".

 

Actually, it's only 0.13C per decade.

 

Oops. Misplaced decimalitiis attack!

 

Actually you didn't read the details of the two as one is for ANOMALIES and the other is ABSOLUTE as well described in the link:

"I’m utilizing the NASA Goddard Institute of Space Studies GISTEMP global dataset. The difference is simply this – I show both the absolute (measured) and the anomaly (statistically magnified) versions of the global temperature. This is accomplished by doing the reverse procedure as outlined in UCAR’s How to Measure Global Average Temperature in Five Easy Steps."

===

The second chart is the true representation of data as it is measured data while the other one is a departure from average statistically generated and plotted which is why that misleading slope is visible.

You should read the whole article that easily explains why they are DIFFERENT and only one of them is honest which you completely missed.

 

I admit I did miss the different calculations you applied to the data, though it's still the same underlying data and I'm not sure you can simply declare either the "anomaly" or "absolute" measures right or wrong. In all cases, you need to understand what the data represents to understand it.

Regardless, the main reason I missed the different calculations is because of your choice to present the second graph on an artificially expanded scale which completely obscures the data. If you want to identify the differences in the two datasets, why not present them on the same scale? Could it be that they're not quite as different as you'd like us to believe? Given your core objection is with presentation of data being used to exaggerate a conclusion (with some justification as it happens), it's more than a little ironic that you seem to be doing the same.

 

Honest Joe writes,

You have it completely backwards since it is straight data placed onto the chart showing absolute data only information, it is the Anomaly chart that is misleading since it is a standard departure from baseline average which is why it creates that slope that so excites warmist/alarmists.

Again, from the link that you keep misunderstanding:

I show both the absolute (measured) and the anomaly (statistically magnified) versions of the global temperature.

 

Your second graph has an upward slope as well though, it's just much harder to see because of the unnecessarily wide scale. I'm not challenging the data, I'm challenging the presentation. Only if you present the two datasets on the same scale can you demonstrate how the different forms of measurement/calculation impact the data.

 

The point is that the slope is negligible.

 

The second graph slope only looks negligible because of the scale. Both slopes appear to run from around 57 degrees to 59 degrees. The point is that the underlying data values for each graph are much more alike than they've been designed to show.

Note that the validity of the different forms of calculations or the practical implication of whatever the actual temperate increase could be are different questions. My issue is purely one of presentation, what the actual data represents doesn't actually make any difference. If you have any data set where every value is between 55 and 60, putting it on a graph with a scale from -20 to 120 is ridiculous (or, of course, intentionally misleading), especially if you're specifically contrasting it with a similar dataset you've presented on a scale from 56 to 59.5.

 

And the first graph slope only looks alarming because of the scale.
One of my favorite examples concerns ice mass loss. Because of the big numbers involved, it's possible to make big headlines out of small losses.
Graphing The Icy Reality
2020 › 03 › 22 › graphing-the-icy-reality
Greenland and Antarctica ice loss accelerating ... Earth's great ice sheets, Greenland and Antarctica ... compare to the total ice mass in the Greenland and Antarctica ice sheets? Their ice volumes are not exactly
total Greenland or Antarctic ice yield alleged amounts of ice gain/loss. ... Perhaps a level of ice loss/gain

The data’s all available on two Excel spreadsheets, freely available here. Figure 3 shows my graph of their data corresponding to the “Antarctica” part of Figure 1:


Figure 3. Cumulative ice mass loss, Antarctica. The photo is penguins on surreal ice.

And Figure 4 shows the corresponding data from Greenland:


Figure 4. Cumulative ice mass loss, Greenland. Note the different vertical scales. Greenland loses more ice than Antarctica.

YIKES! The ice loss looks like it’s falling off of an ice cliff …

So those agree with the IMBIE study, and they are both adequately terrifying.

Having seen that, I thought “how does this compare to the total ice mass in the Greenland and Antarctica ice sheets? Their ice volumes are not exactly known but are on the order of thirty million cubic kilometres in Antarctica and a tenth of that, three million cubic kilometres, in Greenland.

Now, one cubic kilometre is about 0.95 gigatonnes of ice. Using those figures, I added the monthly Greenland ice mass loss shown in Figure 4 to the total mass of the ice in Greenland. That gives me the monthly total amount of Greenland ice. Figure 5 shows that result.


Figure 5. Monthly change in Greenland ice mass as calculated but not graphed by the IMBIE team.

See the blue/black line across the top? Yep, that’s the change in Greenland ice. The net change is so small that you can’t really see it even in a quarter century plus of data. It’s about five-thousandths of one percent (0.005%) of the total Greenland ice mass per year … be still, my beating heart.

And here’s the corresponding plot for Antarctica:


Figure 6. Change in Antarctic ice mass as calculated but not graphed by the IMBIE team.

As before, the blue/black line across the top is indeed the change in the total ice mass of Antarctica. The thing is, all of that terrifying ice loss shown in Figure 3 represents a total loss of three ten-thousandths of one percent (0.0003%) of the Antarctic ice mass per year … lost in the noise.

 

And no graphs/data are more misrepresented than the cherry-picked denier blog ones.

I can still recall the No increase in Temperature February Winters in Kyoto ("therefore global warming doesn't exist") graph delivered with a straight face...Lol

Great thread. Keep it up!

 

The Kyoto graph made no claim beyond what it showed. The claim "therefore global warming doesn't exist" was entirely your contribution.

 

I noticed that you are too lazy to look up the IMBIE data where they show the same results which is already made clear at the bottom of the charts.

HAW HAW HAW HAW HAW, you are that blind?

Since you didn't actually disprove anything your post was dead on arrival.

Great thread. Keep it up!

 

Your graph has a ridiculous, unrealistic scale. The advantage of the first graph is it improves resolution so "small" differences are easily seen. The issue is, what counts as a small difference? You seem to be unconcerned about climate change unless it reaches what, 90 degrees, 120 degrees average global temperatures? lol... And while a couple degrees may not feel like much on its own, it's possible that a small difference could have consequences like extreme weather, melting ice, changes in ecosystems. So it's misleading to show a graph and show "oh look it's just a couple degrees." The concerns about climate change aren't just oh it's a little too hot now.

 

Sure, you can word it either way. The point is that your claim is that the problem is with the way the data is calculated and that it is your way for calculating the data for the second graph that presents a more realistic picture of the situation. On that basis, there is absolutely no justification for changing the scale.

I totally agree with the general principle that the way graphs, and data in general, is presented can have a massive impact of how it's perceived, and that is routinely used by all sorts of people and organisations to make data show whatever it is they want it to show. That can certainly include people promoting the proposed impacts of climate change. My core problem remains that you did exactly the same thing in your OP. You could even have a valid point about how that data has been measured and calculated but any evidence of that is essentially hidden behind your ridiculous second graph. It makes it look like you're trying to trick people, even if you actually weren't.

This would all be relatively easy for you to resolve of course - just generate graphs for both sets of data against both scales. That way we'd all be able to clearly see how much the difference is due to the different calculation methods and how much due to the presentation. Are you willing to do that?

 

The problem behind the problem is that anomaly graphs are much easier to tweak for visual impact.

 

How exactly? As I've said, it doesn't matter what the data values actually represent, extending the scale as widely as you did will mask the results regardless. And it doesn't look like the actual values on the two graphs different all that much.

I thought your issue was around how the data is measured/calculated rather then the presentation of the graphs - you even corrected me when I overlooked that in my first response. So again, why don't you simply present both datasets as graphs using both scales to make any actual issues clear?

 

I don't believe I made any such point. My point has always been about misleading presentations.

 

OK, it was another poster who made that point, though you did like their posts and didn't correct their assertions.

So we're back to my initial point. I totally agree that how data is presented in graphs (and indeed other ways) can be misleading (intentionally or otherwise) but I think for the specific data we're talking about here, your second graph is potentially more misleading than the first. If you wanted to address the (in)significance of the 1 to 2 degree changes the data shows, you can certainly discuss, as any scientific context presenting data should and would do so. Simply presenting the second graph alone doesn't encourage that discussion though, it just seeks to mask the data entirely.

 

I need not agree with every bit of a post I find helpful in general.
The second graph does indeed make the point that the change is insignificant.