Earth Temperature Timeline | Hacker News Reader

Earth Temperature Timeline

xkcd.com

145 pointsr7217 comments

As is common, Mr. Munroe put things into context. I sincerely hope this image is shared thoroughly. This "all in perspective" thing might help a lot more than out of context quotes, and to kill the FuD.

However, I might be overly optimistic in light of the human condition.

This does a great job of putting things in perspective.

This picture is simply wrong.

First: Why is the temperature compared to the 1961-1990 average and not to the range of the timeline (20000 BCE - 2016)?

Second: The measurements/estimates were not made in the intervals marked with dots, they were not equally spaced and not exact as the chart suggests.

Third: It also doesn't take variance into account in the past estimates, yet it does so in the recent measurements (1900-).

Fourth: There is no explanation for the drop of average temperature between years 1000 and 1500, although the change was the same as between 1500 and 2000 and it clearly points out that this is (supposedly) due to human actions.

Fifth: It's funny how at 17500 BCE there's a note stating "Temperatures have been creeping upward" (no explanation why), 10500 BCE "warming resumes" (no explanation why), 7500 BCE "This warm stable period is called 'the holocene stable optimum'" (no explanation why) ... But when it comes to modern times we have explanation for everything (= we, the almighty narcissistic creatures).

And I could go on and on.

1) because there's a steep change since then 2) He addresses this, there could have been faster fluctuation that canceled each other out in measurements longer ago, although unlikely.

1) I think you misunderstood me. I think deviations should be compared to the average of the whole data-set (22016 years) not only a tiny (29 years) subset of it.

2) Exactly, but the point is that the recent data also contains normal fluctuations. It's simply wrong to change the representation of data mid-chart. How would you feel looking at a stock price chart where historical data is displayed as an annual average and it then suddenly continues as an hourly graph? Sure you would think some crazy stuff is going on with the stock price.

Re: 2) This is a semi-valid point. We don't have accurate yearly global temp data from 2,000BC or 20,000BC, so as the data gets older, it is necessarily a smoother and smoother average. Ancient yearly temperature variability is simple not something that we have a detailed view to.

Still, we do have decent variability data for the past 500 years, and the rate of change that we are seeing for the past 50 is truly extraordinary compared to the past 500 years.

Based on the data that we do have, the likelihood that the hockey-stick temperature trend is "normal" variability can now be calculated pretty precisely... confidence levels are now in the 99%-99.9% range. (http://phys.org/news/2014-04-statistical-analysis-natural-wa...)

In short, we can say with better than 99% certainty that...

Yes, this is happening. The globe is now growing hotter at a truly extraordinary rate

My only problem with it is that i have no idea how coarse the dotted lines data is

Reply on news.ycombinator.com