As temperature rises, so does water.

World Watch may appear to some readers to have stumbled into a blind turn with its July/August cover story headlining "The Drying of China." About two weeks after that issue came out, the disastrous flooding of the Yangtze River hit the news. China seemed about as wet as a country could be.

But in fact, there is no contradiction between the drying-up of rivers and water tables and the flash flooding of river valleys. In the past year, 21 countries have experienced both major flooding and drought. Both droughts and floods begin as natural events, but can be worsened by human interference - and by the same interference.

In China, the interference has included extensive deforestation for farming and industrial development. When forest is removed, the soil loses much of its organic matter and becomes harder and less able to absorb water. The farms in dry regions draw underground water for irrigation, while industrial plants draw water for processing and cooling, all of which can cause water tables to fall. The industries' carbon emissions contribute to global warming, which increases the rate of evaporation and heightens the demand for irrigation, which causes the tables to fall still further. Hence the drying.

But warming also triggers weather extremes, including more intense monsoons. When the seasonal rains come, because more water went up in evaporation, more comes down in torrents - though not necessarily in the same place. The rains strike land that has been stripped of its forest cover. Instead of soaking in and replenishing the underground supplies, the water rushes over the surface and as it converges from all over the watershed, it gains force. Hence the flooding.

For many, seeing the connection between droughts and floods can be an important step toward a clearer understanding that ecological issues are central to the future of human life. It seems only two or three years since the time when global warming was viewed by most people as a minor issue - if an issue at all. Some apparently thought it only meant that the weather would just get a bit balmier, or that everyone's lawns would just be a little greener. The coal and oil industries' PR campaigns even managed to convince many people that warming might not be happening at all.

We've come a long way since then. It's now clear not just to the climate scientists who began issuing warnings in 1992, but to a growing portion of the media and public, that human-caused climate change could affect just about everything - including what we eat, where we live, what work we do, how safe we are, what prices we pay, what risks we take, what insurance we have, what illnesses we are vulnerable to. All these things are connected.

That's especially worth keeping in mind in assessing the damage that's been done to China. In our "drying" issue, Lester R. Brown and Brian Halweil noted that China's water problems are leading to food supply difficulties that will affect the whole world. But China's flooding, too, should be of concern for the whole world, because the kinds of disruptions that caused it are more widespread than we may have thought.

One factor in the Yangtze disaster was that 85 percent of the original forest cover in the Yangtze watershed has been removed. A second factor is the region's population; whereas a majority of the major watersheds worldwide have fewer than 30 or 40 people per square kilometer, the Yangtze has 224. The combination of torrential surface-water movement and high human density was devastating.

But the Yangtze is by no means unique in this respect. A new study, co-authored by the Worldwatch Institute's Janet Abramovitz along with Carmen Revenga, Siobhan Murray, and Allen Hammond of the World Resources Institute, examines 145 of the world's major watersheds and calculates the extent to which human activity (deforestation, development, agriculture) has altered the ecology of each. The findings of that study, compiled in the book Watersheds of the World: Ecological Value and Vulnerability, show that 22 of those basins have suffered even higher losses of original forest than the Yangtze.

The Ganges may be of particular concern, because it [TABULAR DATA OMITTED] equals or exceeds the Yangtze's vulnerabilities on three counts: it has the same level of deforestation along with substantially higher population density and a greater number of cities. Significantly, the Ganges delta too was the site of catastrophic flooding during the past few months, as more than half of the densely populated nation of Bangladesh was inundated. Two other Indian watersheds, the Krishna and the Tapti, also exceed the Yangtze in population density. So does Europe's Rhine.

If population density is measured by the number of large dries (over 100,000 population), the Yangtze's count of 56 is exceeded by the Mississippi (87), Ganges (82), Rhine (68) and Danube (60). A table of selected major watersheds, above, suggests that the collision of human population growth, deforestation, and climate change felt in the Yangtze could be felt on every continent. Measures of vulnerability that exceed those of the Yangtze appear in boldface.

Of course, deforestation and population density aren't the only factors determining vulnerability of human communities to floods. Topography is a major factor; if a river comes out of high mountains, chances of heavy snowmelts and fast downhill flow may be greater. (Unfortunately, the Ganges gets no reprieve on that count; it flows down from the highest mountain range on Earth.) And if monsoon rains already provide baseline levels of natural flooding, the exacerbation caused by human interference may be worse. That interference can include, in addition to deforestation, such alterations of the landscape as building dams or dikes, or draining of wetlands.

The point here is not to try to predict where disaster will occur next, but to show that what happened in China could happen in other places - and to varying degrees already has. In 1997 and 1998, according to the U.N. Office for the Coordination of Humanitarian Affairs, there has been disastrous flooding in 54 countries, and the worst could be yet to come. On all continents there are watersheds fed by water falling from high altitude (Alps, Zagros, Rockies, Andes, East African Rift, Himalayas), and on all continents the water increasingly finds dense human populations in its path. If the flow of water is heightened by human disturbance both in the air and on the ground (faster evaporation, heavier rainfall, faster snowmelt, faster runoff, poorer absorption), the results aren't hard to imagine.

What all this says is that what we do about climate involves facing much more than the hardships endured by fossil fuel industries, or by the GDPs they prop up. Failure to sharply reduce carbon emissions will hit people where they live, all over the world. The global climate treaty offers us nothing more than a small beginning, as Christopher Flavin's article in this issue warns. But it's worth keeping in mind that this beginning will very likely lead, if we know what's good for us, to much larger, more revolutionary, changes. We can already see the signs, in the sheer numbers of people being desiccated, drowned, or displaced.

Corrections: Bill McKibben's population projection of 230 million Americans in 2050, as quoted in Curtis Runyan's September/October review of his book, Maybe One, is based on a 50 percent reduction of immigration levels to 400,000 people each year, as well as on the noted drop in the fertility level of U.S. women to 1.5.

Jeff Kenworthy's co-author of the book Cities and Automobile Dependence is Peter Newman, not Peter Hall as stated in "How Mid-Sized Cities Can Avoid Strangulation" (September/October 1998). They have recently released Sustainability and Cities: Overcoming Automobile Dependence (Washington, DC: Island Press, November 1998).