@Terry For those of you that don’t understand what the problem with this is, it’s grabbing a cherrypicked range of data to create a fake trendline. To see the actual trend, you have to look at all the data.
@dave@Terry that stair stepping looks statistically bizarre. a crazy person might infer that they actually change how they measure the data every 10 years or so and it would make it invalid to use all of it together as a single data set.
@PapaPole@Terry No, the stair stepping is showing you how you can grab a subset of the temp data - because in the short term it is fairly noisy, and there are short-term effects such as the El Nino/La Nina cycle that impact temps slightly - and with that create a trendline that isn’t representative of the overall trend.
The problem is this only works for short intervals. Eventually you’re gonna get another warm year that is warmer than all previous years, and the trendline goes positive again. Once that happens they’ll switch to a different data source (there are several programs that try to model global temperature averages, and they’re all going to vary slightly year to year) and start the cycle again with a different dataset.