One of the more awkward facts of California's hydrology is that 70 percent of the state's annual runoff of fresh water occurs north of Sacramento, whereas 80 percent of the state's water consumption takes place south of that city. To supply the south, increasing amounts of water have been diverted from the Sacramento and San Joaquin rivers, greatly reducing freshwater inflows to San Francisco Bay. These diversions have been of great interest to scientists concerned with the health of the bay, which is tied to fluctuations in salinity. They have looked closely at the flow from the Sacramento-San Joaquin Delta, the complex of islands and channels where the two rivers meet, which accounts for 90 percent of the freshwater inflow to the bay.

But sorting out the causes of the water-flow and salinity fluctuations in the delta and San Francisco Bay, and in many other estuary systems, is not a simple matter. A wide variety of climatic and human influences act on the estuary. Fluctuations in climate would cause freshwater inflow to the bay to vary dramatically from year to year, even without large diversions upstream. And the diversions themselves vary from year to year.

It is not surprising, given all these influences, that the salinity of the bay is highly variable and has been rising. Between winter and summer most years, salinity varies as much as 10 parts per thousand--an enormous fluctuation when one considers that the salinity of coastal ocean water is normally just over 33 parts per thousand. And bay salinity varies by a similar amount from year to year. Over the longer term--a matter of particular concern--spring salinities have been slowly rising over the past few decades, increasing by 3 parts per thousand since 1941.

The diversion of fresh water is a large part of the story. Largely because of diversions for agricultural uses, it is estimated that the delta flow is less than 50 percent of its volume in 1850 (although estimates are uncertain because flows were not measured before development took place). Diversion is clearly responsible for much of the salinity increase, but does it account for all of it? To what extent, for instance, might the salinity trend reflect natural fluctuations, such as winter warming or a shift to weather patterns that favor the upwelling of saline water off the coast?

It is one thing to ask these questions and quite another to answer them. Distinguishing short- and long-term anthropogenic trends from the fluctuations of a natural system can be very difficult. To make this distinction in estuarine dynamics we must take a much broader view of estuarine systems than scientists have usually taken.

An example of the approach taken in the past is a sprawling scale model of San Francisco Bay located in a warehouse in Sausalito that has sometimes been used to study the effects of water exports on salinity distributions. When a salinity study is run on the model, the freshwater supply is programmed to vary as it has been observed to vary during field surveys, and the salinity distribution in the scaled-down estuary is then measured with and without proposed exports. Given the way the problem is framed, any change in the salinity distribution that a study uncovers is necessarily caused by freshwater diversions rather than by natural fluctuations.

Like the bay-model studies, most scientific studies of estuarine dynamics have taken the known variability of river flow as a given. At the U.S. Geological Survey, however, we are also attempting to discover why river flow varies. To do so we are studying the estuary as a component of the global climate system rather than in isolation. In effect we have taken the roof off the Sausalito warehouse and knocked down its walls, allowing it to experience the same storms as are supplying its inflow from the high Sierras. By this means, we have been able to demonstrate that much of the year-to-year variability and part of the long-term salinity trend do indeed result from natural fluctuations in the large-scale atmospheric circulation patterns that govern the weather over California.

We suspect that this broader approach to modeling estuarine dynamics will become more common as scientists are increasingly called on to distinguish between the effects of climate and of human activities on estuarine variables. In Chesapeake Bay, for example, one focus of concern is an increase in the volume of anoxic, or oxygen-depleted, water. People have doubtless contributed to the oxygen depletion by allowing sewage and fertilizers, which fuel the growth of oxygen-consuming organisms, to flow into the bay. The spring runoff also contributes to anoxia, however, by stratifying bay water and thus preventing oxygen from being exchanged between the atmosphere and the depths of the bay. It has been suggested that spring runoff may have increased over the past century because of deforestation in the bay's watershed. But deforestation cannot explain year-to-year fluctuations in the volume of anoxic water, which appear to be caused instead by climate fluctuations. To untangle the contributions of these two forcing factors, scientists must include within their model of the estuarine system the whole river basin and the prevailing atmospheric circulation patterns, and not just the estuary itself.

Part 2