Global Warming Was A Fifth Worse In The Last 150 Years Than Historical Records Show


Robin Andrews

Science & Policy Writer

Models appear to show a warmer world than calculations based on historical data. Why is that? lexaarts/Shutterstock

When making predictions, scientists tend to be quite conservative. On occasion, this can mean that they have underestimated something, and as a new comprehensive study has shown, one of these things may be the extent of historic man-made climate change.

Writing in Nature Climate Change, a team from NASA’s Jet Propulsion Laboratory (JPL) points out that there are a few “quirks” in the historical records of global temperature changes, small lapses in data collection that climatologists are acutely aware of. These range from changes in how temperature is measured in some parts of the world to there simply being not enough weather stations to get acceptable amounts of local climate readings.


As this study points out, models that circumvent these quirks, such as those used by the authoritative Intergovernmental Panel on Climate Change (IPCC), show a comparatively warmer future world than about 90 percent of calculations that are based solely on historical records. Some have pointed to these apparently cooler historical record calculations as evidence that predictive models are overestimating the degree of future global temperature rises.

This team of researchers wanted to know if applying these quirks would remove this discrepancy. They performed a series of rigorous calculations on both historical record data and commonly used climate change models, with and without the quirks taken into account, in order to attempt to replicate pre-existing, real-world data.

The team found that as much as 19 percent of air-temperature based global warming over the past 150 years may have been underrepresented in the historical record. They discovered that, when the quirks were taken into account, the historical records matched up far more closely with the modelling studies.

This means that the most recent IPCC modelling studies are absolutely not overestimating temperature changes – they really are showing us what the next few decades hold in terms of global warming.



The Arctic is often underrepresented in terms of its temperature records. British Columbia Ministry of Transport

These so-called quirks tend to revolve around the ever-changing nature of Earth’s water. Historical records mix both air and water temperatures. Water takes longer to heat up than air, so when both are averaged out during studies looking at historical trends, the world seems like it's warming slower than it actually is.

Other quirks are simply sampling problems. The Arctic, for example, is warming up faster than anywhere else on Earth, but due to its inaccessibility, there are fewer historic readings from there compared to lower latitudes. Consequently, a sparser data set seems to show that the Arctic was warming far slower than it is known to be now.

“[These quirks are] quite small on their own, but they add up in the same direction,” Mark Richardson, a researcher at JPL and lead author of the study, said in a statement. “We were surprised that they added up to such a big effect.”


Ultimately, this means that climate change models are more-or-less accurately describing how fast the world will warm in the near-future. This is good news for the scientific method, but bad news for the world at large.

Many models and studies indicate that rises in global temperature will be directly or indirectly responsible for dangerous sea level rise, economic decline, powerful natural disasters, forest cover depletion, ocean acidification, agricultural failure, an unprecedented refugee crisis, and the wholesale destruction of the Arctic. This new study seems to suggest that they are, unfortunately, bang on the money.


Climate change refugees are becoming increasingly common. Piyaset/Shutterstock


  • tag
  • climate change,

  • nasa,

  • global warming,

  • ipcc,

  • masked,

  • historical records,

  • temperature rises,

  • underestimate,

  • overestimate