• Climate

Claim that the Earth has cooled since Medieval or Roman times is contradicted by available data

Posted on:  2019-09-09

Key takeaway

There have been regional climate events over the past 2,000 years, which can be seen in records of past climate. However, analysis of global temperatures for this time period shows that current warming is unique in its extent and has exceeded earlier temperatures.

Reviewed content


Peer-reviewed studies, geologic records, and all the studies have shown that we have actually cooled since the Roman Warming Period, and likely since the Medieval Warming Period.

Source: TFP Student Action, Marc Morano, 2019-08-13

Verdict detail

Factually Inaccurate: Published research actually shows that the last century is the warmest of the last 2,000 years, globally.
Fails to grasp significance of observation: The existence of warmer or cooler periods in individual records from a region does not, on its own, represent a global climate event.

Full Claim

Peer-reviewed studies, geologic records, and all the studies have shown that we have actually cooled since the Roman Warming Period, and likely since the Medieval Warming Period.

A recent study1 analyzing a global database of paleoclimate records found that no previous warm or cool period in the last 2,000 years—including the Roman Warm Period and the Medieval Warm Period (also called the Medieval Climate Anomaly)—occurred globally and synchronously. But 20th Century temperatures were the warmest of the last 2,000 years for nearly the entire surface of the Earth.

figure from study: maps showing timing of peaks of warm or cool periods for each location in the past with color scale
These maps show the timing of the warmest temperatures in named warm periods (or coldest temperatures in named cool periods) over the last 2,000 years. Only the 20th Century warming is global in extent (top right).
Source: Neukom et al (2019)1
Kevin Anchukaitis member picture

Kevin Anchukaitis

Associate Professor, The University of Arizona

1) Our most up-to-date understanding of global mean temperatures is likely the Last Millennium Reanalysis1. Even accounting for uncertainties, global mean annual temperatures are higher now than any time in the last 2,000 years, with the middle of the 20th century either matching or exceeding Common Era temperatures as well.

2) For the Northern Hemisphere (where we have much better proxy data and a better understanding of the uncertainties), we come to similar conclusions—in Wilson et al (2016)2 we find that the two warmest decades of our reconstruction (from 918 to 2004 CE) were 1994–2003 and 1946–1955. Coming in 3rd place is 1161–1170 CE. So, the latest in large-scale temperature reconstructions do NOT support a claim that temperatures either in the Medieval or Roman periods were warmer than today.

3) Even if these periods were warmer than today (and we currently have no evidence that they were), that would have no bearing on whether CO2 is causing current warming. Current warming is unambiguously caused by CO2 emissions. Past warm and cold periods reflect a mix of internal climate system variability and changes in radiative forcing from volcanoes and solar variability. Current warming is the result of rising atmospheric CO2 concentration, whether or not there were past epochs of widespread warming.

Rob Wilson member picture

Rob Wilson

Professor, University of St Andrews

Since 2015, several tree-ring based studies of large-scale Northern Hemisphere summer temperatures have shown that the last 10 or 20 year periods are significantly warmer than any other time for the last 1,200 years. A recent study in Nature1 has expanded on this work using a global multi-proxy data-set which suggests that the last 50 years have been warmer than any period of the last 2,000 years.

The statement above is therefore incorrect and mis-characterises what is detailed in many studies. In fact, basic conclusions have been rather consistent over the past 20 years—i.e., recent warming is unprecedented for the last millennium (likely for last 2,000), the Sun has almost no detectable attributed impact on climate, and volcanoes have the strongest impact on climate prior to the anthropogenic period. Our knowledge of past climate is better where we have data and poor where we rely on teleconnections and/or interpolation.

However, the paleoclimate community and users of the data we generate must appreciate that the 1st millennium of the Common Era is substantially much less constrained than the last 1,000 years and more effort and investment is needed to increase the number of climate proxy records for this earlier period (also the Medieval). Ideally, we should not mix proxy records that express different signals as the climate response to both internal and external forcing varies across different seasons (i.e., summer vs. winter). I would also contentiously add that for the late Holocene, we should minimise the use of proxy archives with poor (>10 years) resolution and substantial (+/- 5-10 years) dating uncertainties.

[These comments are taken from a previous review of a similar claim.]

Science Feedback is a non-partisan, non-profit organization dedicated to science education. Our reviews are crowdsourced directly from a community of scientists with relevant expertise. We strive to explain whether and why information is or is not consistent with the science and to help readers know which news to trust.
Please get in touch if you have any comment or think there is an important claim or article that would need to be reviewed.

Published on:


Related Articles