INTRODUCTION:

Michael E. Mann,
Department of Environmental Sciences, University of Virginia, Charlottesville, VA

SPEAKERS:

Henry N. Pollack,
Professor of Geophysics, Department of Geological Sciences, University of Michigan, Ann Arbor, MI

David Easterling,
Principal Scientist, National Climatic Data Center, National Oceanic and Atmospheric Administration, Asheville, NC

Thomas R. Knutson,
Geophysical Fluid Dynamics Laboratory (GFDL), National Oceanic and Atmospheric Administration, Princeton, NJ

Temperature changes that occur at the Earth's surface propagate slowly downward into the rocks beneath the surface. Thus, rock temperatures at shallow depths provide evidence of changes that have occurred at the surface in the recent past. The pace of heat transfer in rocks is such that the past 500 years of surface temperature history is imprinted on and contained within the upper 500 meters of the Earth's crust.

Analyses of underground temperature measurements from more than six hundred boreholes from all continents except Antarctica show that:

• The global average ground surface temperature has increased by at least 0.9�F (0.5�C) in the 20th century. This is a conservative estimate of the century-long rate of warming because many boreholes used in this study were drilled and logged 15 to 20 years ago, prior to the extraordinary warming of the final decades of the 20th century.
• The 20th century has been the warmest century of the last five centuries.
• The present-day mean temperature is at least 1.8�F (1.0�C) warmer than five centuries ago; of this change about half has occurred in the 20th century alone, and 80% has occurred since the year 1800.

The five-century change can be thought of as a time- and space-averaged overall measure of climate sensitivity (the response of the global mean surface temperature to changes in climate forcing factors over this time interval).

These interpretations provide an historical perspective that indicates that the 20th century has not been just another century in terms of temperature change. In the context of the five-century interval investigated, the 20th century is clearly unusual.

Changes in Temperature Extremes

One of several pieces of evidence used to gauge climate change is an increase in extreme climate events. The two types of extremes examined here are: (1) record-breaking average global temperatures, and (2) changes in the number of days in the U.S. where the temperature exceeds or drops below a given threshold temperature (e.g., freezing).

Evidence from paleoclimatic data suggests that current temperatures are the warmest in the past 1000 years, and more recent observations of global temperatures indicate that temperatures have warmed approximately 0.6 �C (1.1 � F) over the past 100 years. However, an important piece of information related to understanding the sensitivity of the climate system to increases in carbon dioxide and other greenhouse gases, is the rate of warming. Since 1990, society has witnessed some of the warmest years on record. In particular, 1997, 1998, and now 1999, are the three warmest years on record. Furthermore, embedded within the temperature records of 1997 and 1998, was a string of sixteen consecutive months where the monthly global temperature broke the previous record for that month. In fact, during much of 1998, monthly records were broken that had just been set the previous year.

Changes in the Rates of Temperature Change

A casual inspection of the global temperature time series reveals that the increase in global mean temperature has not been constant. A simple linear fit to the time series from 1880 to 1999 shows that there are actually two periods where the rate of change has been much more than the observed 0.6 �C (1.1�F)/100 years, and two periods where the rate of change has been slightly negative or very close to zero. Three inflection points (places where trends change direction) in the above time series were identified using statistical methods, then linear trends were fit to the sub-sections of the time series as defined by the presence of these inflection points. Analysis of these trends show a slight cooling of - 0.38�C (-0.7 �F)/100 years from 1880 to 1910, a strong warming trend of 1.2�C (2.2�F)/100 years from 1911 to the 1940s, a slight cooling of -0.27�C (-0.5�F)/100 years from the 1940s to mid-1970s, and a very strong warming on the order of 0.2�C (0.4�F)/decade since the mid-1970s. Using this information, the string of sixteen consecutive months of record-breaking temperatures was analyzed for consistency with this observed rate of warming over the past two decades. Results of this analysis suggest that this string of record-breaking temperatures in 1997-98, is not consistent with a rate of warming of 0.2�C (0.4 �F)/decade, but may signal an increase in this rate of change. In fact, the observed rate of change since the 1970s is comparable to the 1995 IPCC "business as usual" model scenarios of human-induced climate change for the 21st century which give a rate of warming of about 2.0�C (3.6�F)/100 years.

Changes in Daily and Yearly Temperatures in the U.S.

The average climate warming observed within the continental United States is about 1�F (0.5�C) over the past 100 years. It has been shown that most of the warming represented by the global average temperature is associated more with warming in minimum temperatures (nighttime lows) than in maximum temperatures (daytime highs). Analysis of changes in the number of days where the minimum temperature dips below freezing indicates that, for the U.S. as a whole, there has been a decline of two fewer days per year where temperatures fall below 0�C (32�F). However, since the southeastern U.S. is one of the few places in the world that has exhibited a cooling, there has been an increase in this region in the number of days below freezing. In contrast, the western U.S. has witnessed significant decreases in the number of days below freezing.

The ability of global climate models to reproduce the observed surface temperature trends over the 20th century represents an important test of the models. Confidence in the ability of climate models to anticipate future climate changes rests in part on such evaluations. A recent set of global climate model experiments involving the use of a GFDL model, and driven by past concentrations of greenhouse gases and an estimate of the forcing by anthropogenic sulfate aerosols, was compared with historical temperature observations at various geographic or spatial scales: 1) global mean surface temperatures; 2) latitudinally-averaged temperature changes for various latitudes; and 3) maps of temperature trends for regions of the globe with sufficient observations.

The model used provides a fairly realistic simulation of 20th century surface temperatures in terms of both global averages and latitudinally-averaged temperatures for various latitudes. Comparison of smaller-scale regional details of trends over the last half-century indicates that some significant discrepancies remain between model output and observations. In other words, in some regions, the difference between the model's trend from the greenhouse + sulfate experiments and the observed trend is greater than the "margin of error" as estimated by the internal climate variability in the model. However, for all spatial scales examined (including regional scales) the aggregate model results suggest that these regional warming trends are unlikely to be the result of internal climate variability alone, and suggest a role for a sustained climate forcing resulting from the buildup of greenhouse gases in the 20th century.

• 1. Global Mean Temperature: The model driven only by estimates of the varying concentrations of greenhouse gases and sulfate aerosols over time is capable of reproducing the 20th century observed global-mean surface warming quite well. Five such simulations using different initial ocean states simulate an overall 20th century warming that closely approximates the observed warming.
• 2. Latitudinally-Averaged Temperatures: Observed surface temperature changes, averaged for different latitudes, show that the warming since the 1970s has been fairly uniform across different latitudes. In contrast, the early 20th century warming was largest in high latitudes of the Northern Hemisphere. This difference in spatial structure suggests that the early 20th century warming may have resulted from a different set of causal factors than the recent warming. All five of the model experiments show a warming since the 1970s that is fairly uniform across different latitudes, similar to the observations. This result suggests an interpretation of the late 20th century record as a greenhouse-gas-induced warming "signal" emerging from the background "noise" of internal climate variations. One model experiment suggests that internal climate variability may well have played a substantial role in the early 20th century warming.
• 3. Geographical Pattern of Trends: The most stringent of the tests applied is the comparison of the complete spatial pattern of the observed and simulated temperature trends. For the model to be in agreement with observations, it must agree not only in terms of the globally averaged temperature changes, but also in terms of the regional details. According to this test, the climate model forced by greenhouse gases and sulfate aerosols is statistically consistent with the observed trends over more than 2/3 of the globe (considering only regions with sufficient observations). However, significant discrepancies exists between model and observations over about 30% of the area examined. Nonetheless, the observed warming trends appear to be clearly outside the range of internal climate variability alone.

In the case of the greenhouse + sulfate experiments, a number of factors may contribute to the regional discrepancies between observed and simulated trend patterns. These factors (ordered in terms of our estimate as to their relative importance, from most to least important), include possible deficiencies in:1) specified radiative forcings such as indirect sulfate aerosol effects, for example; 2) climate model sensitivity to the forcings; 3) simulated internal climate variability, especially regionally; or 4) observational records.

Dr. Henry Pollack is a professor of geophysics in the Department of Geological Sciences at the University of Michigan in Ann Arbor. He has engaged in research on all seven continents, addressing the dynamics and evolution of the Earth and its climate. His current research focuses on the record of global climate change as recorded by the temperatures of the rocks beneath the Earth's surface.

Dr. Pollack has served on National Science Foundation advisory panels on Continental Dynamics, the Global Digital Seismograph Network, the San Andreas Fault, and Earth Science Instrumentation and Facilities. From 1991-95, he served as Chairman of the International Heat Flow Commission of the International Association of Seismology and Physics of the Earth's Interior. He is presently a member of the U.S. Geodynamics Committee of the National Research Council, and the Committee on Global and Environmental Change of the American Geophysical Union.

Dr. Pollack received his undergraduate degree in geology from Cornell University, an M.S. degree from the University of Nebraska, and a Ph.D. in geophysics from the University of Michigan. He has also held visiting teaching and/or research positions at Harvard University, the University of Zambia, the Universities of Durham and Newcastle (UK), and the University of Western Ontario.

Dr. David Easterling is currently Principal Scientist at the National Climatic Data Center (NCDC) in Asheville, NC. Prior to that he served as an Assistant Professor in Climate and Meteorology, in the Department of Geography at Indiana University-Bloomington. He has authored or co-authored numerous research articles in such peer-reviewed journals as Science and the Journal of Climate. He is also a contributor to the upcoming Intergovernmental Panel on Climate Change (IPCC) 3rd Assessment Report, and serves as a member of the National Assessment Synthesis Team.

Dr. Easterling's research interests include the detection of climate change in the observed record, particularly changes in extreme climate events; the development of statistical methods for improving the quality of climate data; and the application of General Circulation Model simulations in developing climate change scenarios for use in assessing the potential effects of climate change on the environment and society. He received his Ph.D. from the University of North Carolina at Chapel Hill in 1987.

Thomas Knutson is a research meteorologist in the Climate Dynamics Group at the National Oceanic and Atmospheric Administration's Geophysical Fluid Dynamics Laboratory, one of the world's leading climate modeling centers. He has been author or co-author of a number of publications in major climate research journals, including two recent papers in Science on future hurricane intensities under a global warming, and on a model simulation of early 20th century global warming. His recent research interests include: detection of climate change; simulation of internal climate variability; and the impact of climate change on El Niño and hurricanes. He has been an invited expert on climate change and extreme events at the Aspen Global Change Institute, the Risk Prediction Initiative of the Atlantic Global Change Institute at the Bermuda Biological Station for Research, and at a recent Environmental Protection Agency workshop on climate change and extreme storm events. More recently, he was an invited speaker at the American Meteorological Society's 23rd Conference on Hurricanes and Tropical Meteorology.