Cumulative days over Global Temperature Milestones

Following my frustrations with Climate Reporting – Why so many different values, and the 6 individual days where GMST was above 2C (E.g. above the upper limit of the Paris Agreement) … I was starting to wonder, for how long we would have experienced these high temperatures before the “GMST Trend Temperature” was agreed to have passed these temperature milestones. E.g. when will it be before the Paris Agreement 1.5 / 2.0 Limits are agreed to have passed.

This graphic above (Beware: log scale), clearly shows when the world first experiences a single day above a global temperature milestone, and shows how rapidly that new temperature milestone becomes a regular occurrence. Within a few years, after “GMST Trend Temperature” passes a milestone, the curves become very smooth, because virtually every day after that is over that milestone.

Graph – 2nd Version October 2024

The second version of this graph was created using Copernicus ERA-5 daily GMST temperatures up to 1st October 2024.

• 612 individual days above 1.50C Paris Agreement global temperature target
• “GMST Trend Temperature” is well on its way to 1.5C
• 100 individual days over 1.75C (Note: no change here, from April 2024)
• 6 individual days over 2.00C Paris Agreement global temperature target
• Improvements in second version
  • Trend line
    • The first version talked of a year when “climate reached” a milestone, and this used a home-made smoothing function, which was a mix of centre-moving-average and linear-best-fit.
    • The second version used standard “Loess Smoothing” with a 30-year-window, to calculate the years which the “Loess Smoothing Trend Line” passed various milestones.
    • This actually made surprisingly little difference.
      • 1st version stated that we passed 0.5C in 1984. This second version shows Loess Smoothing Trendline passes 0.5C in 1985.
      • 1st version stated that we passed 1.25C in 2020. This second version shows Loess Smoothing Trendline passes 1.25C in 2021.
  • Counts
    • As per Copernicus 1850-1900 Baseline – Daily GMST Anomaly, I did a deep dive into some differences between the numbers I got for Copernicus data relative 1850-1900, versus those published by Copernicus themselves (in their monthly bulletins)
    • The differences were only around 0.03C for any given day.
    • The improvements mean the counts for cumulative days over 0.5C, 0.75C, 1.00C, 1.25C were changed by about 0.2%. The counts for 1.5C, 1.75C, 2.0C were not affected.
  • Twitter / X feedback
    • I had some really useful criticism / feedback from the graphic’s first posting on Twitter / X.

Graph – 1st Version – April 2024

The first version of this graph was created using Copernicus ERA-5 daily GMST temperatures up to 2nd April 2024.

• 506 individual days above 1.50C Paris Agreement global temperature target
• “GMST Trend Temperature” is well on its way to 1.5C
• 100 individual days over 1.75C
• 6 individual days over 2.00C Paris Agreement global temperature target
• Note that the Y-Axis is Logarithmic.

I wanted to tie together the accumulating individual days above various GMST Milestones, to “The Climate”. This is overreach by me. For future postings of the graphic (with the latest data), I will use a more anodyne “GMST Trend °C” … E.g. when do my trend lines pass various “GMST Milestones”, being clear about what trend line algorithm I use.

Key Data Points Show Accelerated Warming

The table below appears to show accelerating warming:

• Note:
  • The table was updated in Oct-2024, as I moved to Loess Smoothing with a 30-year-window
  • This page only uses Copernicus ERA5 data.
• The years between “GMST Milestones” keeps dropping:
  • 15 years from 0.5C to 0.75C, but only 9 years from 1.00C to 1.25C
• The cumulative number of days to reach a “GMST Milestone” keeps dropping:
  • Well Over 2,000 individual days above 0.5C before “GMST Trend Temperature” reaches 0.5C
  • Just over 1,000 individual days above 1.25C before “GMST Trend Temperature” reaches 1.25C

How the Graphic was constructed

The graphic uses the “Copernicus ERA-5” daily Global Mean Surface Temperatures (GMST) , together with “Daily Anomalies” (See GMST Data Sets). This allowed me to get the Daily GMST above the 1850-1900 baseline.

For each day between January 1940 and April 2024, the graph simply counts the cumulative number of days that were above 0.5C, 0.75C, 1.00C, 1.25C, 1.50C, 1.75C, 2.00C. The Y-Axis is Logarithmic, so that it can fit the 14,000+ days over 0.5C on the same graph as the 6 days over 2.0C.

The Year at which the “GMST Anomaly Trend Temperature has Reached <X>°C”, is somewhat subjective, however using various smoothing techniques that give believable “GMST Anomaly Trend Temperature” lines, it only most only makes 0-2 years difference for any given data set. Currently (Oct 2024 onwards) I am using Loess-Smoothing with a 30-year-window on the Copernicus data.

• Climate Milestones – NOAA, GISS, Copernicus, HadCRUT, Berkeley Earth
  • Determining the “GMST Trend Temperature” using multiple GMST Data Sets and smoothing functions
• Climate Milestones Copernicus – ERA5
  • Determining the “GMST Trend Temperature” using just the Copernicus Data set (See GMST Data Sets) and two different smoothing functions
• As per Climate Reporting – Why so many different values
  • If you did the same graphic, but with different data sets (E.g. one with “Berkeley Earth” and another with NOAA), you could get up to 7 years difference in the dates, as their “trend / smoothed” data varies by up to 0.17C

Page History

5-October-2024 – Updated to Loess Smoothing

As per Log of Updates and Errata updates in Oct 2024, I moved away from a hand-woven smoothing technique, and moved towards using Loess Smoothing with a 30-year-data window. This didn’t make a huge difference, but is a popular smoothing technique.

7-April-2024 – Updates in progress…

Following feedback on Twitter / Blog – I am making a few changes

• Replaced terms:
  • “The Climate” with “GMST Trend Temperature”
  • “Climate Milestones” with “GMST milestones”
  • It was pointed out that “The Climate” is not the best term. I guess it is also “overreach” on my part. I don’t want to distract from the main message of the graphic, so choosing a less “punchy” title / terminology.
  • I’m expecting to post updates of the graphic every month, so the key title will just be a count of the days over various GMST milestones.
• Adding “+/-” range to the “year that the “GMST Trend Temperature” crossed a “GMST-milestone”
  • This is to reflect the subjective nature of these years and the trendlines.
  • I need to find a better trend-line-algorithm. Maybe “Loess”
  • Some people lost confidence in the graphic, because it shows such an early crossing of 1.25C GMST-Milestone. The trend lines are above the year-average temperatures for 2021 / 2022. This is actually ok, because 2021/22 were La Nina years and you would expect some years to be above, and others below a trend line.
• Better title and y-axis name
• Add an additional label on the right, about the the “days > 0.5” .. being cumulative since 1940
  • Some people thought it was “x days over 0.5C, since trendline passed 0.5C”
  • … whereas it is actually “x individual days above 0.5C since Jan 1st 1940”
• Added a better graph to Climate Milestones Copernicus – ERA5, to show more clearly how the “TrendLines” weave their way between individual yearly averages. Some years (E.g. 2021 / 2022) will be below the Trendline, and others (notably: 2016, 2023, start-of-2024) will be above the trend line.
• Improvements still needed
  • Need better colours. Need to find a standard palette for GMST Milestones