We write to address a serious mischaracterization in your article about our scientific results on the evidence for changes in temperature and climate during the last millennium (News, “Warming Study Draws Fire,” Sept. 12).
Your article implies that we tried to determine the globally-averaged temperature going back 1,000 years. We did not, owing to limitations in the records of indicators of environmental change. The global temperature can only be reliably reconstructed using records of surface thermometers for the past two centuries. That global record of surface temperatures shows the 20th century to be warmer than the 19th century.
The focus of our research was on climate, which is more than just temperature, over many centuries, well beyond the range of widespread thermometer records. As has been customary for over three decades in studies of past climate, we relied on environmental indicators of change, including annually-dated information in tree rings, coral, stalagmites and ice cores. Those indicators respond differently to temperature, rainfall, sea salinity, etc., so they cannot be simply averaged. The indicators are not sampled across most of the world to give a global picture.
Our papers thoroughly document the difficulties in estimating past global temperatures from an amalgamation of such disparate environmental indicators. We concluded that it is not possible at present to construct an accurate millennial global temperature record. Even our sharpest critics agree.
Willie H. Soon and Sallie L. Baliunas
Sept. 25, 2003
The authors are scientists at the Harvard-Smithsonian Center for Astrophysics.