I posted
two days ago the TempLS results for July, showing that is was 0.225°C hotter than any prvious July (2019), following on from a record hot June. These measures tipped the results to date above the previous hottest years of 2016 and 2020. But whereas those years peaked in February/March, it seems that 2023 is just getting started. So I have made some graphs to show where things stand.
The first is one of a kind that I tracked the progress of 2015 and 2016. It shows the year to date averages of recent hot years over the twelve months. It isn't ideal for 2023, because the early warmth of 2016, say, is exaggerated. I have marked with a fine grey line the previous record, which 2023 to date has just passed. But more significant are the black dots of the actual 2023 monthly temperatures. It means that if anything like the July level can be maintained, 2023 will be, so to speak, a boilover.
Next is a graphical version of the
table of ordered months that I wrote about last month and added to the data page. It shows a rectangle for each of the warmer months, the top of which is the temperature for thet month,and the bottom is that of the next lower month. The annual average is to the right (year to date for 2023). The key for year colors is at the top.
It emphasises by how much June and July 2023 exceeded previous months. This could be seen as overstated, since other record-setting years have been overwritten by more recent hot months. So here now is a plot of the months as they would have shown when they set the record, ie the years descending are monotonic. That shows how the years of 2015 and 2016 were more remarkable at the time. But again, I think 2023 is just getting started.
Finally I'll show an updated plot of the time history of hot year records. It is just a regular annual plot, but with colors to show the years for which each record lasted, showing also the rapid recent rise.