Taking the Planet’s Temperature: 2026 Edition

This is a follow-up to my article from January, 2024, titled: “How Hot is Hell? I mean Earth.”

Of course, it is common knowledge that 2024 was the hottest year in recorded history and likely the hottest year in over 120,000 years, at about 1.6°C above the 1850-1900 baseline (Copernicus data). If the long-term average was one year alone, we could comfortably say the Paris limit has been broken. But that’s not climate, that’s one very hot year due to a variety of transient impacts whose future contribution is uncertain.

Where are we today? What is the “global temperature anomaly with respect to the 1850-1900 IPCC pre-industrial baseline as of January 1, 2026? That’s the question this article will answer.

The most recent data point I can find is from a January 14, 2026 article in The Guardian, which states:

Current rates of heating could breach the Paris Agreement limit of 1.5°C (2.7°F) – which is measured over 30 years to iron out natural fluctuations – before the end of the decade, according to the EU’s Copernicus climate agency.

Take note that the Paris Agreement does not specify a methodology for computing the long-term global temperature anomaly. The Paris Agreement does not state 20 years or 30 years. It certainly does not state a “trailing 30 year average” as implied in this quote. Also, by saying that the Paris limit is expected to be breached before the end of the decade, this quote implies that the current status of the anomaly is significantly less than 1.5°C. As we will see, I disagree.

It’s nuts that the “anomaly” was not defined in the text of the Paris Agreement. It’s nuts that for over a decade, people have been making up non-existent ways this number is officially computed.

So, how do we measure the long-term temperature in a reasonable way? This article by Richard Betts et. al. from 2023 explains the methodology that I will employ. Here is the pertinent part:

In other words, the temperature should be based on a 20-year average, but not the trailing 20 years. Rather, it should be the average of the previous 10 years as actually measured, together with the forecast for the next 10 years based on models, with the current day (January 1, 2026) in the middle of that span. Average all those temperatures together and that’s where we are today.

I decided to do this computation using this definition together with data from Copernicus ERA5 (available here). This data gives accurate daily global surface temperatures for the period 1940-2025. Another file available from Copernicus gives their 1850-1900 pre-industrial baseline, day-by-day. The data in these two files allowed me to compute the day-to-day anomaly with respect to the pre-industrial baseline for every single day from 1940 to the present.

The key step is to focus on “decadal average gains.” For this, I considered each day from January 1, 1975 to December 31, 2025, and compared that day’s global mean surface temperature with the same day from 10 years earlier. Focusing on a single day gives a lot of variability, so I included a “least squares” linear trendline to get a sense for the average gain per decade. I then extended the trendline 10 years into the future, to get the decadal gain through the end of 2035. Here’s what I got:

If you examine this graph closely, you will see that as of the end of 2025 we are roughly at 0.32°C gain per decade. Because the line is tilted upwards, the rate of decadal growth is steadily increasing. More precisely, each year the decadal warming appears to rise by roughly 0.0044°C over the rate from the start of the previous year. This is the very definition of accelerated warming, which is what Hansen et. al. have been claiming for a long time. For example see this article in The Guardian from November, 2023.

To estimate the current global temperature using Betts’ method, I used this trendline. I simply took the equation for the trendline and used that to forecast the daily temperatures a decade into the future. Each day from January 1, 2026 through December 31, 2035 was forecast by taking the temperature from a decade earlier and adding the appropriate increase that the trendline gave for that day. Averaging those values with the actual recorded temperatures from January 1, 2016 through December 31, 2025 gave the answer.

Assuming accelerating warming, the global temperature anomaly with respect to the 1850-1900 IPCC pre-industrial baseline as of January 1, 2026 is 1.45°C.

Also, given the current rate of warming, the forecast anomaly for January 1, 2027 is 1.48°C and for January 1, 2028 it’s at 1.51°C. So, as of this writing, with accelerating warming, the Paris limit will be breached sometime in 2027.

In summary:

• At the start of 2026, we are 1.45°C above the 1850-1900 pre-industrial IPCC baseline

• The current decadal average gain is 0.32°C and rising.

• Global temperatures are accelerating.

• We will break the Paris limit by the end of 2027.

Now, to be clear, this is not a statement about what I expect to happen in 2026. The temperature in any given year is dependent on many different transient effects and 2026 is already showing the potential for an El Nino to develop later in the year. If this comes to pass, the average for 2026 may very well approach 1.60°C. It’s important to keep the long-term average separate from the short-term volatility that various climate oscillations produce.

Take it or not, I am satisfied with this result. It has the look, feel and smell of being right. And isn’t that what climate science is really all about anyway?