society to optimize its activities and prepare for the changes in the most cost-effective way possible. These capabilities will enhance our environmental security and sustain our continued economic success.
This chapter expands on some of the important, most societally relevant influences exerted by these attributes, documents our knowledge of how they have changed in the past, and explains why we have chosen to focus on them now.
Freshwater is the very basis of terrestrial life, and is arguably its most precious natural resource. Water influences nearly every aspect of society and day-to-day life. From the huge amounts of freshwater required for modem industrial and agricultural production to the bucket of clean water so highly prized in less developed countries, the uninterrupted supply of clean freshwater is necessary for the overall health and continuance of our societies and economies. In addition, freshwater distribution influences energy production and utilization, water quality, fisheries and land ecosystems, forestry, insurance, recreation, and transportation. The longevity of aquifers, the reliable flow of rivers, and the fall of rain determine where civilizations can grow and prosper. Significant investments in our infrastructure, such as the construction of dams and levees, and water-resource planning and management in general, are based on our current understanding of the supply, storage, and dispersal of freshwater. Any changes or disruptions in the freshwater cycle as we have come to know and rely on it can thus have widespread consequences, with implications for all levels of society and every individual in it.
Variations in the water supply will have more serious effects on some societies than on others. Less developed countries, particularly those with semi-add climates, marginal agriculture, and rigid social structures, are clearly vulnerable to growing-season failures: The history of northeastern Brazil is replete with examples of major failures of growing-season rainfall (see Figure 2-1) that caused mass migrations of Nordestinos to other parts of Brazil (Magalhaes and Magee, 1994). More developed societies, through their economic prowess, are less vulnerable to the year-to-year variations of precipitation. For example, the record rainfalls over the midsection of the United States in June to August of 1993 led to record flooding (Kunkel et al., 1994; Bell and Janowiak, 1995); the Mississippi River was above flood stage for almost three months at St. Louis. This resulted in extraordinary damage (estimated at $15-20 billionsee Changnon, 1996), yet the flood, while causing considerable local hardship, produced only a blip in the U.S. economy. Similarly, the record summer drought of 1988 caused an estimated $30 billion in agricultural damage alone (Trenberth and Branstator, 1992), but the strength of the U.S. economy (if not the balance sheets of the people in the region) was easily able to withstand this climatic event.
As the time scale of precipitation variability increases, even the most developed countries become vulnerable. The United States enjoys an enviable agricultural sector that has become more efficient over the years, employing an ever-decreasing share of the population in the task of feeding its, and the world' s, people, and showing great resilience in recovering from the random flood or drought, no matter how severe. But during the 1930s, when the economy was particularly fragile, an entire decade of low rainfall caused migrations and dislocations in the United States similar to those of northeastern Brazil. Indeed, recent paleoclimatic evidence from enclosed lakes (Laird et al., 1996) suggests that such droughts were considerably more severe and longer-lived in the past, relative to what we have experienced in the past few hundred years (Figure 2-2). Other parts of the paleoclimate record also suggest that such severe droughts are not unprecedented. Even today, when the economy of the United States is far more stable than during the Depression years, a decade of poor rainfall in the fertile agricultural regions would lead to economic dislocations and would place grave strains on the national and global economy. More frequent occurrence of floods like those in the Midwest in 1993 and 1997 would have similar types of effects.
The patterns of rainfall in the Sahel region of Africa also show decadal- and centennial-scale variability (Figure 2-3). The devastating impact of prolonged low rainfall on the mostly nomadic societies of sub-Saharan Africa has required massive and continuing infusions of world resources to avoid even greater disasters (Glantz, 1994). Such long periods of drought leave little room for adaptation by vulnerable populations, or for future economic development, and they affect the intellectual development of children in ways that will echo through generations.
An even longer drought may well have spelled the doom of the Classic Maya civilization (Hodell et al., 1995). The