If we look at the record of the last century, we see that U.S. precipitation has increased by 5-10 percent. To put this in perspective, this increase is equivalent to the annual amount of water that flows out of the Mississippi River.

This increase has a lot of regional texture. Some states are suffering from lack of rain, but in many parts of the U.S., the increases have been several times greater than the average. In 1996 alone, six states set "all-time high" records for precipitation. Furthermore, since the beginning of this century, the frequency of intense precipitation events, where more than two inches of rain falls in a 24-hour period, have increased by about 20 percent. Such events lead to flooding, soil erosion, and even loss of life.

We know only too well how costly weather disasters can be. Since 1980 we have had 30 weather or climate events with price tags of at least $1 billion. Nineteen of the events have occurred in the last 5 years. Among the most costly were Hurricane Andrew in 1992 ($30 billion); the Mid-West floods of 1993 ($20 billion); 1995's hail storms and floods in Texas, Oklahoma, Louisiana, and Mississippi ($5.5 billion); the Southern plains drought of 1996 ($4 billion); and Hurricane Fran in 1996 ($5 billion). The winter storms that struck the Pacific Northwest in December 1996 and January 1997 are now estimated to have resulted in over $3 billion in costs and the estimated cost of Northern Plains floods that occurred in April and May of 1997 is about $2 billion and still rising. California is being ravaged by flooding this week from El Niño-associated storms with more to come.

We are seeing dramatic evidence of the costs associated with climate variability as the Western U.S. deals with this year's unprecedented El Niño event. I want to be clear that we can't say that El Niño is itself a consequence of climate change. We are gratified that we now have tools to predict El Niño conditions up to a year in advance, but I hasten to add that we still do not understand what causes such oscillations in the first place. However, we do have good long-term records and better instrumental records of this phenomena in modern times, and these indicate that the behavior of El Niños in the last twenty years or so has been quite anomalous. The disruption of what was a more regular El Niño cycle has been coincident with the recent pronounced warming of the Earth surface temperatures, and some scientist believe there may be a relationship, but no-one has produced any conclusive evidence. We have made great progress in explaining the effect of El Niño on the global climate, and our skill at predicting these events is improving markedly (e.g. we predicted nearly a year ago the Pacific storms we are now experiencing, and that Atlantic hurricanes would be diminished during this period. Yet even advance warning that makes it possible to undertake advance mitigation measures, the impacts are significant. Damage estimates in California are already very high, and we expect this total to grow over the next several months. El Niño provides a kind of case study of the vulnerability of our society to climate disruption, and should serve as a warning to us of the potential consequences of climate change.

The costs of such extreme events have doubled or tripled in each of the last few decades. This is partly because our urban and technologically advanced society has become extremely dependent on a massively integrated infrastructure for power, communications, transportation

and fresh and wastewater treatment and distribution. This infrastructure is extremely vulnerable to extreme weather events. I want to emphasize that one can't point to any single extreme weather event today and say for sure that global warming caused it. But we can say that such events are examples of the kinds of impacts we expect to occur with greater frequency in a warmer world. There are likely to be more "storms of the century," "100-year floods," and severe droughts in the future than there were in the past.

The temperature increases, intensification of the water cycle, and sea-level rise already observed over the past century are all consistent with theoretical predictions of the consequences of an enhanced greenhouse effect. They are also consistent with the projections from simulations of global climate by general circulation models.

The IPCC "business as usual" scenario indicates that even with continued technological improvement (such as energy efficiency increases of about 1 percent per year), uniess policies to control emissions of greenhouse gases are implemented, the atmospheric concentrations of these gases will be much higher by 2100. Assuming "business as usual" CO, concentrations will reach about 710 parts per million by volume (ppm), a level higher than any seen on this planet in the last 50 million years (Figure 2). For context, the pre-industrial level of CO, was about 280 ppm, and has increased to the current level of about 360 ppm. If realized, this increase is expected to result in significant future climate changes:

Global surface temperature would increase an average of another 2-6 °F by 2100, with a best estimate of 3.5 °F. Higher Northern latitudes are projected to warm by more. Temperature change of this magnitude would be faster than any observed changes in the last 10,000 years.

Global mean sea level would rise another 6 to 38 inches by the end of the 21st century. The rate of evaporation would increase as the climate warms, leading to an increase in average global precipitation as well as frequency of intense rainfall and floods in some regions. In some regions, the soil moisture will decrease, leading to increased frequency and intensity of droughts.

Most of the climate impacts have been evaluated for a world at equilibrium after greenhouse gases have reached either 550 or 700 ppm. But stabilizing at double the preindustrial concentration of greenhouse gases, or 550 ppm, would require massive intervention. On the other hand, a continuation of "business as usual" implies a world with far higher concentrations and far greater effects. The Geophysical Fluid Dynamics Laboratory at Princeton has recently modeled the effects of doubling and quadrupling the level of greenhouse gases:

A quadrupling of such concentrations (to about 1100 ppm) is likely to increase
temperatures in North America by 15 - 20° F, as opposed to the 5-10° F expected from
doubling.

In the growing season, soil moisture deficits would approach 30 - 50 percent for
quadrupling, as opposed to 10 - 30 percent for doubling.

The implications of this amount of greenhouse gases in the atmosphere are only dimly perceived, but would be grave.

Although there is scientific uncertainty about exactly how and when the Earth's climate will respond to increased concentrations of greenhouse gases in the future, observations clearly show that detectable changes are underway. Numerous efforts are underway to assess the vulnerability of ecological, social and economic systems to future changes by defining plausible future scenarios of climate change and using model simulations to evaluate sensitivities to various levels and types of change.

The Current Situation

I find the scientific evidence for climate change in the coming decades to be compelling, and the implications for the future, without intervention, potentially disastrous. I believe that we need to confront this growing challenge now. It is clear that emissions of greenhouse gases from human activities are amplifying the Earth's natural greenhouse effect, and are leading to a warming of the planet's surface. It is also clear that this warming will, in turn, lead to a series of further climate disruptions as sea-levels rise, patterns of precipitation change, atmospheric and potentially ocean currents shift, and ideal ranges for plants and animals change faster than nature can accommodate them. Climate change is a long-term challenge, and solving this problem will require a sustained, long-term effort. Thoughtful reaction to lessen the undesirable impacts of climate change can simultaneously alleviate other problems such as air pollution and growing dependence on fossil fuels.

The longer we continue "business as usual," the faster the rate of climate change, the greater the degree of warming, and the more severe the negative effects for human and ecological systems. The sooner we leave the "business as usual" path, the sooner we can slow the rate of change and avoid more negative impacts, and the more likely it is that natural systems will be able to adapt to change, and the less probable such unanticipated events as ocean current instabilities linked both to El Niño and Northern European climate.

Climate change will force shifts in the range and distribution of many individual plant and animal species, and thus will alter many of the ecosystems, including forests and wetlands, that provide the support systems for all life on Earth. The present systems, like the present human infrastructure, evolved and optimized over about 10,000 years of relatively constant climate. If CO, concentration rises to 700 parts per million, as I believe it almost certainly will if we do not take action, fully one third of forests worldwide are likely to experience shifts in species composition. As a small example, in the U.S., sugar maples and beech trees may move completely into Canada, with considerable economic impact. The tourism, maple syrup production, and wood production which are so important to New England's livelihood are at risk. Ideal ranges for crops will change and agricultural pests may increase. Species that cannot migrate fast or far enough may face extinction.

We spend $2 - 3 billion dollars annually to operate and maintain our parks and refuges as unique assemblages of plants and animals. Climate change threatens this substantial national investment. Coastal flooding and storm damages may increase due to rising sea levels. One third of Florida's Everglades may be inundated, and much of the Louisiana and Mississippi coastal lands may be lost. In some Western and Northeastern areas, 50 percent or more of brown trout may be lost as waters warm too much to support them. Malaria mosquitoes may be able to survive year-round in much of the U.S., and drought may become more frequent even as total rainfall and extreme precipitation events increase. Functions provided by the natural environment, such as flood control and purification of air and water, can only be provided by intact ecological systems, and when those systems are disrupted or lost we lose the enormous benefits they provide.

The potential cost of such disruption is very serious. Several recent analyses have addressed the issue of valuation of ecosystems. Refining and extending this initial work is a necessity, as was just reinforced by a report by the President's Committee of Advisors on Science and Technology, "Teaming with Life: Investing in Science to Understand and Use America's Living Capital." Initial estimates of the total value of global ecosystem services suggest they could be in the trillions of dollars annually.

For ecosystems, the rate of temperature change may be even more significant than its eventual magnitude. As the distinguished ecologist Professor Jane Lubchenco told the President earlier this year: "The slower the rate of change in climate, the less catastrophic the results. Species are more likely to be able to migrate, to grow, and evolve if the rate of change is slow." The sooner we reduce emissions, the slower the rate of climate change we can expect.

The Framework Convention on Climate Change, to which the United States is a party, seeks to stabilize atmospheric concentrations at levels that "prevent dangerous anthropogenic interference with the climate system. Such a level should be achieved within a time frame sufficient to: (1) allow ecosystems to adapt naturally to climate change; (ii) ensure that food production is not threatened; and (iii) enable economic development to proceed in a sustainable manner." Greenhouse gas concentration are already beyond the range seen in the last 160,000 years, and one could argue that the rate of climate change is already outside that range as well. Even if emissions were kept at today's levels, temperatures would rise at about 2 degrees F per century because greenhouse gas concentrations would continue to rise. The historical pollen record shows that this rate is about the fastest ecosystems migrate.

Atmospheric concentrations of CO, have increased thirty percent since the onset of the industrial revolution. Analyses that model global economic conditions and technological change suggest that to hold atmospheric concentrations to 550 ppm of CO2, which represents about a doubling of pre-industrial levels, reductions in the growth rate of worldwide annual emissions must begin by about 2015. That is, if we want to follow a least-cost, long-term adjustment path, it appears that the whole world must begin to "apply the brakes" within the next 2 decades, and

global emissions must begin declining steeply during the second quarter of the next century (Figure 3).

The results that give us cause for concern come from numerous studies of past climate change, from observations of ecosystems changing over time, and from model simulations that project future conditions. An important aspect of interpreting such results, especially model simulations, is accounting for uncertainty, which is often done in a very selective way by climate change skeptics. Uncertainty cuts both ways: outcomes could be less dramatic than expected based on our current understanding, but could just as well be more severe. We must remember that the current generation of impact projections is focused on a limited number of variables, and that this may result in an overly benign picture of the future.

For instance, a series of projections that includes carbon dioxide increases, temperature change, and precipitation changes now indicate that while there might be no "net" effect on global agricultural productivity, significant regional dislocations are expected, with the poorest countries experiencing the greatest losses. However, these analyses do not include potential water constraints or changes in the distribution of pests. As the scientists told us at the White House Conference on Climate Change last October 6, the impacts we have not been able to include are more likely to make the net results worse rather than better.

Further, climate change may not occur gradually. Most climate model simulations to date have assumed that climate will change relatively steadily as a result of human-induced emissions of greenhouse gases. Over the past few decades, however, there have been some indications that this might not be the case -- that instead climate might change in abrupt jumps.

Past climate conditions reconstructed from ice core records in Greenland, from ocean sediment cores in the neighboring oceanic regions, and from pollen records on land areas bordering the North Atlantic have found that climates of the past have changed dramatically over very short periods of time -- namely, over a few decades or even less. We have discovered that the apparent mechanism for the climate shifts is change in the large-scale ocean circulation of the North Atlantic (which is coupled to the circulation for much of the world). The evidence shows that in the past, the Gulf Stream has basically stopped carrying heat northward, causing significant cooling in Europe and northeastern North America. The risk of such a catastrophe exists but has been little studied; similarly the potential for abrupt sea level rise from collapse of the West Antarctic ice sheet or runaway warming from dramatic heating of the polar regions and subsequent release of methane cannot be ruled out.

If we do not alter our emissions trajectory, we will pass atmospheric concentrations of double the pre-industrial level by the middle of the next century, on the way toward a tripling, or even quadrupling. We could very well take the planet to levels that have not been seen for 50 million years in a single century, a geological blink of an eye. The Administration has received