Methane (ppb)

have risen from fairly steady background levels (~270ppmv) to present day levels (370ppmv) in a little over a century. This rate of change has no parallel in the historical past, just as temperatures recorded in the late 20th century were unprecedented. Most of the change in CO2 and other greenhouse gases resulted from the growth of world population and the insatiable demand for fossil fuel-based energy. Given that world population will almost certainly double within the lifetime of those currently in kindergarten, unless something is done to curb the use of fossil fuel consumption, it seems very likely that significant changes in climate will occur in the near future.

Figure 2. Changes in atmospheric carbon dioxide and methane levels in the atmosphere over the last 420,000 years (from gas bubbles trapped in an ice core, from Vostok, Antarctica).

Should we be concerned that the climate may change significantly in the future? Here I have focused exclusively on changes in temperature, but temperature change is only one component of our overall climate system. Changes in temperature are associated with variations in rainfall and the amounts of snow, shifts in storm tracks and hurricanes etc. From the record of past climate, we know that a relatively small overall change in global temperature can have significant environmental effects. The "Little Ice Age" was characterized by dramatic changes in ice cover in mountain regions throughout the world. But historical records from lowland areas of Europe also document more extensive snow cover, longer periods when rivers and lakes were frozen over and frequent cold, wet summers, with disastrous consequences for agriculture, leading to social disruption and political upheavals. Such changes all occurred with an overall change in average hemispheric temperature of less than 1°F. Of course, in trying to anticipate the effects of future climate change, we are looking at the consequences of warmer, not colder conditions but the implication is the same-even a small shift in average global or hemispheric temperature, with its associated changes in atmospheric circulation, rainfall patterns etc., can be highly disruptive to society. We have seen many examples of such anomalies in recent decades, yet temperatures, though warm, were nowhere near the levels that may be reached later in this century. These include extremes in rainfall, leading to catastrophic flooding in some areas, and droughts, exceptional wildfires and historically low lake levels elsewhere, as well as an increase in windstorms and other weather-related disasters. Unusual weather events are becoming less uncommon, impacting agriculture, transportation and commercial activity. Of course, such disasters have always occurred to some extent, but the frequency of extremes has increased in recent years throughout the world, leading major insurance companies to express grave concerns about their exposure to these unprecedented risks (note

that these risks are in addition to the costs due to increased development). Munich Re, one of the world's largest re-insurance firms recently reported:

"1999 fits exactly into the long-term pattern of increasing losses from natural catastrophes. insured losses came to $22bn. This is the second highest figure ever recorded windstorms were responsible for 80% of the insured losses while earthquakes accounted for 10%, floods 6%, and other events like forest fires, frost, and heat waves around 4% . . . In view of the fact that the signs of climate change and all its related effects are becoming more and more discern

ible. if. meteorological extremes like torrential rain, windstorms, and

heat waves continue to increase and the rise in sea level accelerates, many regions of the world will be in immediate danger . . ."

Can we be certain that future climate will involve unprecedented risks? Some argue that processes within the climate system will act to compensate for the effects of higher greenhouse gas levels (so-called negative feedback effects). According to this scenario, these feedbacks will help maintain the climatic status quo enabling us to continue to contaminate the atmosphere ad infinitum. There is a small chance that such critics are right, in which case it would be safe to do nothing. But they may be completely wrong, and indeed the scientific consensus is that they are wrong. Political decisions inevitably involve assessing risk and weighing the consequences of action versus inaction. Just as Congress must decide if the (perhaps small) risk of a rogue nation launching a nuclear missile at the United States (resulting in a catastrophe) is worth avoiding by spending large sums of money on a space defense system, so it must weigh the potentially catastrophic environmental and commercial consequences of future global warming against the costs of curbing fossil fuel consumption to reduce these risks. Scientists cannot provide Congress with a certain forecast of the future and as research on global warming continues, our understanding will undoubtedly change. But the picture at present is that we are indeed living in climatically unusual times, and that the future is likely to be even more unusual.

Appendix 1.

Tree ring data include both ring width and ring density variations. Records are available from all continental areas (except Antarctica) though most series are from outside the tropical regions. High latitude and high altitude trees generally provide estimates of past temperature; trees in dry regions generally provide estimates of past precipitation, though even in wetter areas, records of rainfall changes can sometimes be obtained.

Ice cores provide many records of past climate but changes in oxygen isotopes in the ice, accumulation rate and (summer) melt conditions are of primary interest in examining recent centuries. In polar regions oxygen isotopes are generally considered to be an indicator of annual temperature. Other useful climate indicators include the fraction of a core containing 'melt features' (produced by the re-freezing of percolating surface melt water) which provides a useful index of summer temperature conditions, and accumulation rate changes, which indicate past snowfall amounts.

Corals provide uniquely detailed records of sea-surface temperatures, from changes in the (temperature-dependent) oxygen isotopes in the carbonate skeletons of the corals. In some cases, salinity variation is the most important factor influencing isotope content, in which case the changes reflect precipitation and runoff from adjacent continental regions.

Varved sediments, from both lake and marine environments, are annual layers that record past environmental conditions in the lake or oceanic region. There are few ocean areas where varved sediments are known to occur (generally upwelling coastal regions where there is little oxygen in the deep waters) but varved lake sediments are found on all continents. Providing the records are clearly annual and a strong climatic signal can be demonstrated, these records can provide useful data from many regions of the world.

Historical records can, potentially, provide seasonal estimates of past climate over wide geographic regions, though at present only European and East Asian sources have been adequately studied.

Details of how these and other paleoclimate proxies are used to reconstruct past climates can be found in the book "Paleoclimatology" by R.S. Bradley (1999, Academic Press).

The CHAIRMAN. Dr. Christy, welcome.

STATEMENT OF DR. JOHN R. CHRISTY, DIRECTOR,

EARTH SYSTEM SCIENCE CENTER, UNIVERSITY OF ALABAMA Dr. CHRISTY. Thank you, Mr. Chairman. I am pleased to be here testifying before this Committee.

By the way, I am from the University of Alabama in Huntsville. We do not have football team. Ice hockey, in fact, is our favorite sport.

[Laughter.]

Dr. CHRISTY. Considering the varying levels of skepticism represented on this panel, it would be apparent that I am very likely the witness that is most skeptical, but not agnostic, regarding our ability to predict future climate. And I hope to demonstrate why this is so.

The universal feature of climate model projections of global temperature changes due to greenhouse gas increases is a rise in the temperature of the atmosphere from the surface to about 30,000 feet.

This temperature rise itself is projected to be significant at the surface, with increasing magnitude as one rises in the atmosphere, which we call the troposphere.

Over the past 21-years various calculations of surface temperature, indeed, show a rise between .45 and .65 of a degree. This represents about half of the total rise since the end of the 19th Century.

In the troposphere, however, various estimates, which include satellite data that Dr. Roy Spencer of NASA and I produced, show only a very slight warming between .09 and .18 of a degree, a rate less than one-third that observed at the surface.

So rather than seeing a rise in temperature that increases with altitude as climate models project, we see that in the real world since 1979, the rise decreases substantially with altitude.

The most recent modeling efforts which attempt to explain this disparity suggest that when some of the actual climate processes are factored in, and I emphasize "some," such as the Mount Pinatubo eruption, the models looked like they came close to reality.

On closer inspection of these studies, however, one finds that the apparent agreement was achieved only by comparing apples with oranges. The model experiments included some major processes, but not all major processes.

When those additional processes were included, like real El Niños, the climate models did not produce the observed global average vertical temperature changes. In other words, 60 percent of the atmosphere is going in a direction not predicted by models.

And that, in my view, is a significant missing piece of the climate puzzle that introduces considerable uncertainty of the models' utility regarding predicting temperatures.

Now, it is certainly possible that the inability of the climate models to predict what happened over the past 21 years may only indicate that the climate experiences large natural fluctuations in the vertical temperature structure.

However, this means that any attention drawn to the surface temperature rise for the past two decades must, I repeat must, also

acknowledge the fact that 60 percent of the atmospheric mass that was projected to warm did not.

This vertical temperature situation is a curious and unexplained issue regarding global average temperatures. But we do not live 30,000 feet in the atmosphere, and we do not live in a global average. We live in a specific place, city, state, and so on.

Local and regional projections of climate are very difficult and challenging. An example from North Alabama that I wanted to use here, only illustrates the difficulty in providing regional estimates of what might happen.

A few climate models have attempted to reproduce the temperature changes over the last 150 years, since the 19th century. These are complex models with solar changes, carbon dioxide increases, sulfate pollution, oceans, and so on.

They indicate that since the 1890's we in North Alabama should have experienced a warming of about two degrees.

Observations show we have actually experienced a cooling of over two degrees. The models may have done fairly well at the global average surface temperature, and may have done acceptably well in several geographic locations, but my opinion in the southeast, is that there was false information there. I am not hitting climate models in a critical way. I am showing the challenge that is there on reproducing climate results on a regional basis.

If in trying to reproduce the past we see such model errors, one must assume that predicting the future would produce similar opportunity for regional errors. I want to encourage the Committee to be suspicious of media reports in which weather extremes are given as proof of human-induced climate change.

Weather extremes occur somewhere all the time. For example, you have seen recent reports perhaps about the U.S. surface temperature data showing January through March the highest ever in one surface temperature data set of the United States, not others.

The satellite data provides information for the entire globe and show that, yes, indeed, the tropospheric temperatures were well above average for the 48 contiguous states. However, most of the globe experienced below average temperatures in that massive bulk of the atmosphere.

It was our turn to be warm while in places such as the equatorial oceans and the Sahara Desert, it was their turn to be cool. Other climate data give us similar information. Hurricanes have not increased. Tornadoes have not increased. Droughts and wet spells have not statistically increased, or decreased.

Let me quickly add, there are many more people and much more wealth in the paths of these destructive events, so losses have increased but that is not due to climate change. Deaths in U.S. cities are no longer correlated with high temperatures, though deaths still increase during cold temperatures.

When looking at data such as these, especially on a regional basis, climate change, and in particular, the human factor of climate change, is very difficult to detect at all.

I will close with three questions and a plea. Is the climate changing? Yes, it always has and it always will, but it is very difficult to detect on decadal time scales.

Are climate models useful? Yes, and improving. At this point, their utility is mostly in global average scale, yet there are still some significant shortcomings even there.

Is that portion of climate change due to human factors good, bad, or inconsequential? And that, no one knows, although we do know that the plant world thrives on additional CO2 in the atmosphere. What I do know is that we depend on data to answer these questions. The global data network is decaying at the very time we need it most.

If the richest country in the world could do anything, it would be to step up efforts to monitor the present global climate, reconstruct the past climate, assure easy and timely access to data, and to support scientists to study the data on which to depend such important answers.

Thank you.

The CHAIRMAN. Thank you very much, Dr. Christy.

[The prepared statement of Dr. Christy follows:]

PREPARED STATEMENT OF DR. JOHN R. CHRISTY, DIRECTOR,
EARTH SYSTEM SCIENCE CENTER, UNIVERSITY OF ALABAMA

Mr. Chairman and Committee Members, I am pleased to accept your invitation to offer information on climate change along with my own assessment. I am John Christy, Professor of Atmospheric Science and Director of the Earth System Science Center at the University of Alabama in Huntsville.

Carbon Dioxide

The concentration of carbon dioxide (CO2) is increasing in the atmosphere due primarily to the combustion of fossil fuels. It is our great fortune (because we produce so much of it) that CO2 is not a pollutant. In simple terms, CO2 is plant food. The green world we see around us would disappear if not for atmospheric CO2. These plants largely evolved at a time when the atmospheric CO2 concentration was many times what it is today. Indeed, numerous studies indicate the present biosphere is being invigorated by the human-induced rise of CO2. In and of itself, therefore, the increasing concentration of CO2 does not pose a toxic risk to the planet. It is the secondary impact of CO2 that may present challenges to human life in the future. It has been proposed that CO2 increases could cause climate change of a magnitude beyond what naturally occurs that would force costly adaptation or significant ecological stress. For example, sea level rise and/or reduced rainfall would be two possible effects likely to be costly to those regions so affected. Data from the past and projections from climate models are employed to provide insight on these concerns. Climate Models

Climate models attempt to describe the ocean/atmospheric system with equations which approximate the processes of nature. No model is perfect because the system is incredibly complex. One modest goal of model simulations is to describe and predict the evolution of the ocean/atmospheric system in a way that is useful to discover possible environmental hazards which lie ahead. The goal is not to achieve a perfect forecast for every type of weather in every unique geographic region, but to provide information on changes in large-scale features. If in testing models for current large-scale features one finds conflict with observations, this suggests that at least some fundamental process, for example heat transfer, are not adequately described in the models.

Global Averages

A universal feature of climate model projections of global average temperature changes due to enhanced greenhouse gasses is a rise in the temperature of the atmosphere from the surface to 30,000 feet. This temperature rise itself is projected to be significant at the surface, with increasing magnitude as one rises through this layer called the troposphere. Most people use the term Global Warming to describe this temperature rise.