A recent study published in Nature uses paleoclimate records from the 1500s to show that industrial-era warming first became apparent in the Northern Hemisphere in the mid-1800s. This paper suggests that humanity’s climate influence can be seen earlier than previously thought, so current models may underestimate the magnitude of human-caused climate change.

Paleoclimate data is reconstructed by combining observed climate data with known geochemical or biological markers of temperature. This information is analyzed using a statistical model that allows scientists to estimate temperatures for unobserved time points. For example, scientists can combine information from tree rings for years that have temperature readings to learn about how the temperatures affect tree growth. They can then use tree rings to make inferences about the climate in time periods before we started recording climate data, based on the changes that they see in the rings.

This approach can be used with other data sources to model either land or oceanic temperatures. The oceans are a major heat reservoir and play an important role in the speed and structure of climate change, so it is important to understand how human activity may have influenced oceanic temperatures.

The authors of this paper use ten different climate models to see if they can reproduce the paleoclimate data. This approach has been done before with the instrumental record, and it’s clear that the models can’t reproduce the rising temperatures we’ve seen if they don’t consider human carbon dioxide emissions.

Agreement among these different models was particularly pronounced for the Northern Hemisphere regions, where sustained and significant warming began in the northern tropical oceans around the 1830s. This is roughly the same time as increases become apparent in the overall Northern Hemisphere temperature.

In other words, the Northern Hemisphere models and data clearly show the mid-1800s as the time when both the oceans and land masses experienced temperature increases that went beyond prior variability. The models also indicate that this change was due to the effects of industrial-age.

The scientists found it challenging to effectively model Southern Hemisphere temperatures, because there is less historical data available for the south. The data that is available suggests that the onset of industrial warming was approximately 50 years later in Australasia and South America compared to the Northern Hemisphere. However, these estimates are based on limited information and are not as precise as those in the Northern Hemisphere.

Climatologists have previously used models to identify the emergence of climate change, but they’ve primarily used data from the twentieth century to estimate the baseline climate and humanity's influence. The authors write that if the data from this new paleoclimate analysis is accurate, then these previous studies have used temperatures from after the onset of climate change as their pre-climate-change baseline climate. As a result, their models may underplay humanity’s contribution.

The authors suggest that previous studies that attempted to identify the year when climate change became apparent may have underestimated how quickly the effects of climate change forced the Earth’s temperature outside the range of normal climate variability

Sourch;Nature, 2016