A recent post at Watts Up With That? (WUWT) makes a big deal about the fact the National Oceanic and Atmospheric Administration (NOAA) temperature record for the United States changes over time.

This isn’t a new thing. People have known about it for quite some time. It’s a weird effect of the methodology the NOAA uses. It’s wonky. The methodology probably shouldn’t be used. However, it’s nowhere near as bad as a lot of people say it is. People like to claim the changes always cool the past and warm the present, saying things like:

All of the error serves to cool the past, and to warm the present – in other words, all of the error serves to bolster the very assertion being promoted.


They have, in essence (whether they realise it or not), built in a positive feedback loop, the effect of which is to continually cool the past

Anthony Watts, proprietor of the blog, even responded to the first of those saying:

REPLY: I never said it was random, you inserted that idea. I said it was likely due to incompetence, and confirmation bias tends to push that one direction – Anthony

This is ironic as when I compared the two data sets available to me via that WUWT post, I found the NOAA has recently warmed past temperatures while cooling present ones. Don’t believe me? That’s okay. Watts provided the recent temperature data he used. He also provided a link to an earlier post he wrote, which lists the same data as of ~18 months before. This is what you get when you plot the changes between the old data and new:


The WUWT post doesn’t show these changes are the exact opposite of what the people there expect. I get changing estimates of past temperatures can be troubling regardless of what those changes are, but surely the fact recent changes downplay global warming merits some attention. Even if it didn’t need to be pointed out in the main post, it certainly did when directly responding to the person who said:

All of the error serves to cool the past, and to warm the present – in other words, all of the error serves to bolster the very assertion being promoted.

As a word of caution, Watts and I are looking at different data sets. We both use the same, recent data, but he compares it to an older version than I did. He plotted a graph showing temperatures as listed in 2012, but I used temperatures as listed in January, 2013. That’s because I couldn’t find a copy of the data as of 2012.

It would obviously be worth doing similar comparisons for each version. There may have been some changes of the sort commenters at WUWT expect. It’s just ironic people can look at changes which downplay global warming and interpret them as exaggerating global warming.



  1. What you seem to miss, is that I didn’t plot these temperatures at all. The two graphs in this post:

    Were actually produced by NOAA, from their data, with no influence from me whatsoever. I simply annotated the graphs that NOAA created from their climate at a glance plotter. I used in in 2012, and I used it yesterday.

    As proof of this, note that both graphs have NOAA logo watermarks in them, done for the purpose of proving they are a NOAA product.

    You are correct in that these are different datasets at different times. In late 2013 NOAA made a shift to a new dataset that put more emphasis on climate divisions. Unfortunately, they don’t provide the older datasets to the public.

    But the issue here is what they are presenting to the public in their SOTC and their CAAG plotter page. I have only pointed out that they present data differently at different times. I have no control over the inner workings of the CAAG plotter nor any insight to what method it is using.

    Your graph is at odds with what NOAA produced on their CAAG plotter page at two different points in time, I had nothing to do with it.

  2. Anthony Watts, I’m not sure why you think I seem to miss that point. It’s pretty much impossible to miss the fact those graphs originated with NOAA. That doesn’t change the fact versions of the same data set, versions you’ve used, show the exact opposite of what a significant number of your readers expect.

    The graph I created is not at odds with anything. You highlighted changes between two versions of a certain data set. A number of your readers expressed the view changes to that data set always exaggerate global warming. The graph I provided shows that’s not true as changes in that data set (at least sometimes) downplay global warming.

    If I had access to the data used for the first graph, I’d have shown the differences between all three versions. Even without that data though, it seems we can tell the change you highlight downplays global warming. The first graph you show has a warming trend of .124 degrees per decade. The second graph has a warming trend of .1 degrees per decade. Some of that must be caused by the data for 2013 being lower than the data for 2012, but that one point couldn’t cause a change in trend of .24 degrees per decade. Some of it must have been caused by the past being warmed relative to the present.

    In other words, you’ve shown two graphs which highlight a change by the NOAA that downplays global warming, and a number of your readers interpreted it as showing the NOAA exaggerates global warming.

  3. “As a word of caution, Watts and I are looking at different data sets. We both use the same, recent data, but he compares it to an older version than I did. ”

    You say “data sets”. I didn’t actually use the datasets, only OUTPUT from NOAA datasets using NOAA plotter. I did no comparisons of data, did no plots of my own, but you did. In your sentence as I read it, you made it seem as if I plotted those two graphs myself, when I didn’t. I just want the distinction to be clear, mainly because there are some pedants arguing right now at Stoat about a graph Monckton supposedly made, but didn’t, and it was produced by the art department of the Telegraph.

    And today, those same output graphs from NOAA I cite are in a national media article, attributed to me, and of course I already got an email where some idiot claims I produced them “fraudulently” Some people don’t see the watermark if monitor settings don’t show low brightness colors well.

  4. Anthony, the point of providing a graph is for people to be able to see the data in the graph. The point of comparing two graphs is to compare the data in the two graphs. As such, it is reasonable to describe comparing graphs of data as comparing data.

    The odd thing is you stopped your quote right before a statement which I just realized supports your point:

    He plotted a graph showing temperatures as listed in 2012

    I mistakenly said you “plotted a graph” instead of “posted a graph.” I originally wrote that phrase as, “I plotted a graph,” and when I changed “I” to “He,” I forgot to change “plotted” to “posted.”

    I didn’t catch that mistake because it doesn’t change anything about my post. Whether or not you created the images had no bearing on anything I was saying. I tend not to proofread unimportant statements like that as much. So yeah, my bad on that. Sorry. I don’t see that it has much relevance though.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s