The initial view shows three series:

1. The change in the observed 10-year average of the global average surface temperature, relative to that from 1850-1900 (proxy for the climate in pre-industrial times). Observed values are measured directly and fluctuate quite a lot each year due to natural processes (e.g. solar variation, volcanic eruptions, and El Niño oscillations). Averaging over 10 years reduces these fluctuations.

2. The component of the above that is calculated as human-induced; that is, the natural components are mathematically stripped out, leaving behind the warming caused by humans due to greenhouse gas emissions and other forcings (aerosol radiation interaction, aerosol-cloud interaction, black carbon on snow, contrails, ozone, stratospheric H₂O, and land use). Human-induced values change more predictably and smoothly each year than observed values.

3. The uncertainty in the human-induced warming. Note that the observed value has an uncertainty too (typically narrower), but this is not shown.

The human-induced 10-year average can be used to check progress against the long-term temperature goals of the Paris Agreement – the lower 1.5°C goal is shown on the chart, as is an approximate breach year given the current trajectory of human-induced warming, as suggested by the source, see below. The Paris Agreement itself does not specify a method for deciding if certain levels have been reached.

By clicking on the legend you can enable the 1-year average equivalents of the series described above. These show what has happened recently with the climate, which is useful in its own right, though any single year’s observed warming is less useful as an indicator of long-term climate change. However, because human-induced warming is roughly linear, and assuming this continues, the 1-year and 10-year human-induced averages are similar as indicators of long-term climate change. Extrapolating either leads to a first breach of 1.5°C at roughly the same time in the early 2030s. The breach becomes more significant the longer it is sustained.

Note that the observed 1-year average data for 2024 is a rolling 12-month average produced before the end of 2024, hence is not directly comparable to all other data points.

The source, see below, explains all the above in more detail.

To see more clearly how the series are changing from year to year, select ‘Toggle change mode’ in the menu at the top right of the chart. Here you can also change to the temperature projection mode, which is described in the ‘Projection Quick Start’ section.

Graphs can be zoomed by dragging horizontally on the chart (or separate two fingers on touch screens). On desktop, pan left and right when zoomed in using the shift key.