> Since the movie “Inconvenient Truth” the amount of ice of the world has increased significantly toward the mean.
Let's go to that link (or read the top reply) and verify ourselves![0] Should be easy, right? Let's click "hide all years" on the side and also uncheck the median and ranges (so we have nothing). Then one by one, let's click the decade averages in the top right. Where does the graph move? We could also click every 5 years, or even 1 year and watch each line be drawn. Is there a trend? The one I see completely disagrees with that tweet.
But why is the tweet misleading? That's a better question. Well, click 2022 and you'll see something interesting. In Jan-April there's actually a bit more ice than in 2006, with about 0.5 million km2 as a max differential. But July through December it goes far below, with a differential of about a million km2. But let's check other years, to see if there's a bias. Just select 2022, we'll add 2023 and down to 2017. Okay, we see 2022, 2017, and 2020 are a bit different, so let's keep those in mind. Now let's do the same thing around 2006. We see here that 2006 is a bit abnormal for the years surrounding it. Now let's select the bounding years and then compare the two. Does this make sense and agree with the averages? It should.
Your alarm bells should have gone off in the first place because 2023 doesn't even have a full year to plot. Not even half of one.
> Additionally nearly every climate sensor is located on the tarmac of an airport.
Yeah! That's actually exactly right! We need a lot of thermometers and placed in a lot of locations. (I should also mention that there are a lot of thermometer systems that aren't based on airports though but we could also get deep into wet bulb vs dry bulb and all kinds of things. We'll just stay here for now) So we look for existing ones to work with and then use that data. What I find interesting about this tweet is that usually we're criticized for "manipulating" data rather than this. There's a surprisingly good and approachable article on all this here[1]. You'll find how we adjust surface temperature data to account for being near a black body, how ocean temperatures are adjusted based on the measurement techniques that are used (because each one introduces a different bias), and so on.
Actually one thing you'll frequently find me complaining about is how difficult analysis actually is (though usually in the context of ML). This is quite similar. You can't ever really work with "raw" data, because your measurements have biases in them and if you don't account for them you'll get biased results which do not align with observations. But the best first check that you can actually do is look at prediction model and then look at the results and see how good the predictions were. Luckily with something like climate science we've had decades of this going on and so we can verify this pretty well. We just need to look at models created in the 80's, 90's, and 00's and see how good they predicted the subsequent decades of temperature rise[2,3,4].
So I can get the misunderstandings here and the reason to be skeptical. But I think we've thoroughly shown that the two tweets you linked to are either by people that are misunderstanding a highly complex thing (pretty understandable) or are not acting in good faith (which if they're claiming expertise with lacking knowledge, this too is not in good faith). Climate is a pretty complicated topic with a lot of moving parts that can appear to contradict one another if you aren't careful. I highly suggest drawing charts to keep track of how things connect. Most people aren't trained in these things and so it is understandable. But you also should focus on asking questions when you don't understand rather than asserting "facts". There's no reason to feel embarrassed for not knowing things. We all are pretty fucking stupid, myself included. But let's also try to not trick ourselves. The best way to do that seems to be to challenge our own beliefs. You set up a challenge of "how would I disprove myself" and then see if you can do so. If you can't, then congrats, your opinion should get stronger. If you do, dang, let's update our beliefs. Usually it is somewhere in between though and you don't have to tear everything down. But I have to admit, when the whole perception changes, it is kinda exhilarating. Finding out new things is quite a lot of fun.
[0] https://nsidc.org/arcticseaicenews/charctic-interactive-sea-...
[1] https://arstechnica.com/science/2016/01/thorough-not-thoroug...
[2] https://climate.nasa.gov/news/2943/study-confirms-climate-mo...
[3] https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/201...
[4] https://www.columbia.edu/~jeh1/mailings/2020/20200203_Models...