10 responses to “Newborn

  1. All the best for the baby, the parents and the uncle!

  2. Well-said, with all the painful sincerity of the young. This is why climate change is an issue of generational justice. Climate change will probably not kill me–and even if it does, I will have gotten to enjoy a natural world that was largely functional for the bulk of my lifetime.

    That won’t be the case for our young speaker, much less for his niece. Now we’re playing to hold the line at a ‘more or less functional’ natural world.

    Play hard, my friends.

  3. An OT post, but perhaps an idea worth considering. Via an ongoing conversation with a denialist, I came to the idea of comparing the last 41/50 years of GMST with the first 41/50. I say “41/50” because for various reasons I was using a *retrospective* 10-year mean as a metric, but ended up comparing both the averaged and raw annual anomalies. Thus:

    ‘Modern’ data:

    Raw: 1975 to 2016 (41 data points)
    10-year retrospective: 1966-75 mean to 2007-16 mean (41 data points, but labeled to reflect the whole spans influencing the means–ie., 1850-1900 and 1966-2016.)

    (And similarly for ‘early’ data.)

    Here’s the resulting graph:

    That gave rise to some considerations.

    Obviously, the biggest difference is the trend in the modern data. It’ll be interesting to see what defense mechanism my denialist friend deploys in that regard–probably ‘repression’ (ie., simply ignoring it.)

    But another difference is that the early data seem to be more variable. My first thought was that perhaps the trend was creating autocorrelation in the modern data and the result was greater variability in the early data, since there’s little to no trend there. But without really thinking through the logic of that in depth, I observed that the difference seemed to exist in the raw data, too, so that explanation didn’t really fly. My next thought was that the data quality is probably much lower for the early data, so perhaps a feature of that would be greater susceptibility to random variation. Any more informed thoughts on that, anyone?

    But the biggest question I had was, if you have more statistical sophistication than I do–and most participants here do, I fear!–then does anything interesting drop out of the comparison between early and lata HadCRUT4 data as if they were separate data sets? I know just enough to know that there are many, many ways to potentially address that question. So I’m punting to those who may have some chops in that regard, in the spirit of hope and curiosity.

  4. Alternate link to graph, in case the first one doesn’t work:

    (Please delete and/or substitute as appropriate.)

    • Note: I got curious about how many year-on-year declines in the decadal retrospective value there were for each curve. The answers are that for the ‘early’ curve, 24 of 40 comparisons showed a decline (cooling), which is pretty much what you’d expect for a nearly trendless period, but that the ‘modern’ curve shows only 6 of 40 years featuring a decadal decline. There’s only one instance where you get two consecutive declines: it’s in 2011-12, toward the end of the so-called ‘pause’ years.

      It’s pretty much a truism that warming was never expected to be monotonic, but it gets surprisingly close by this metric.

  5. Happy New Year to all and a heads up:
    This from Sou on A Kevin Cowtan paper looks interesting!

  6. Thanks, tokodave, that *is* interesting!

  7. Who is this speaking in the video?

    [Response: He’s Adam Levine, and regularly posts climate-related videos on youtube under the moniker “ClimateAdam” — look him up on youtube.]