Introduction
The ability of resource agencies to manage fish, wildlife and freshwater supplies of San Francisco Bay estuary requires an integrated knowledge of the relations between the biota and their physical environment. A key factor in these relations is the role of salinity in determining both the physical and the biological character of the estuary. The saltiness of the water, and particularly its seasonal and interannual patterns of variability, affects which aquatic species live where within the estuary. Salinity also determines where water can and cannot be diverted for human consumption and irrigated agriculture, and plays a role in determining the capacity of the estuary to cleanse itself of wastes. In short, salinity is a fundamental property of estuarine physics and chemistry that, in turn, determines the biological characteristics of each estuary.
Freshwater is a major control on estuarine salinity. Most freshwater supplied to the Bay is from river flow through the Delta, which is primarily runoff from the Sierra Nevada. Most contaminants in San Francisco Bay are from the Sacramento/San Joaquin Valley and the local watershed around the Bay rather than the sea or atmosphere. Land is the primary source of freshwater and freshwater serves as a tracer of land-derived substances such as the trace metals (copper, lead and selenium), pesticides and plant nutrients (nitrate and phosphate).
The U.S. Geological Survey is collaborating with other agencies and institutions in studying San Francisco Bay salinity using field observations and numerical simulations to define the physical processes that control salinity. The issues that arise from salinity fluctuations, however, differ in the northern and southern parts of the bay. In North Bay we need to know how salinity responds to freshwater flow through the Sacramento/San Joaquin Delta; this knowledge will benefit water managers who determine how much delta flow is needed a) to protect freshwater supplies for municipal water use and b) modulate salinity for a healthy estuary. In South Bay we need to know where the freshwater comes from (the distant Delta or local streams) to sort out the sources of a) contamination or b) dilution.
Salinity Indicates the State of the Estuary
Salinity is a measure of how much sea salt is contained in a unit of water. California coastal seawater is about 33 parts sea salt per thousand parts water by weight (defined in practical salinity units, or psu). Salinity of freshwater is near zero. Thus, for an estuarine salinity of 11 psu, the water is approximately two-thirds fresh and one-third seawater. Salinity of the ocean is relatively constant. Therefore, most of the salinity variations in the estuary are responses to river inflow and mixing by the ocean tides.
Just as body temperature is a state variable of human health, estuarine salinity is emerging as a key “state variable” for fisheries management in San Francisco Bay. Peak spring delta flows and their fluctuations are important for maintaining appropriate habitat for fish species. Because the Bay’s fisheries resources have been sharply declining over the last two decades, resource managers are interested in how salinity varies in spring especially in the interior region of the estuary near the delta, a region that is considered to be a nursery area for important species. Water and fisheries managers need to know how natural and human-induced salinity changes will affect both water availability and fisheries especially during periods of low spring and summer/fall river inflow. Understanding the nature of salinity patterns in the estuary is also necessary for interpreting the movement of toxic substances and/or essential nutrients.
The causes of salinity variations in major estuaries like San Francisco Bay are well-known. Typically, the influences from direct precipitation, evaporation, and leakage from ground water systems are small. Instead, most of the variations in salinity both in space and time are caused by 1) patterns of freshwater discharge from tributary rivers and 2) mixing of freshwater with seawater by both tidal action and wind-driven wave action.
The flow of freshwater into the Bay is largely from runoff generated by precipitation from winter storms and snowmelt carried by the Sacramento and San Joaquin Rivers, a pattern typical of a Mediterranean climate (wet winters and dry summers). During California’s dry seasons, summer and fall, saltwater from the Pacific Ocean moves landward within San Francisco Bay; during the wet winter season saltwater retreats seaward, driven by the increased discharge of freshwater. In wet winters, saltwater is pushed farther seaward (and has farther to return in summer) than in dry winters.
In discussing the nature and causes of salinity variations, San Francisco Bay is commonly divided into two parts, the North Bay and the South Bay (fig. 1). A major issue in the North Bay is the quantification of the salinity response to “delta discharge”: that portion of the flow of freshwater from the Sacramento and San Joaquin drainage basins that passes through the delta into San Francisco Bay. Excluded from delta discharge is the freshwater that is consumed upstream, within the delta, or exported via the pumps of the state and federal water projects.
![San Francisco Bay estuary, with locations of sampling sites](images/sal_fig1.gif)
Figure 1. San Francisco Bay estuary, with locations of sampling sites.
In South San Francisco Bay, a major issue is the need to distinguish the influence of local stream discharge and waste water input from those of delta discharge on water chemistry and contaminant levels.
Importance of Salinity to Resources in North San Francisco Bay
In North San Francisco Bay, delta discharge is the major control on seasonal to yearly variations in salinity. Freshwater inflow also affects the day-to-day fluctuations in salinity, but on these shorter time scales tides and winds also play a role. Most of the controls on delta discharge are ultimately found, in turn, in large-scale atmospheric circulation over the Pacific Ocean, which during the course of its west to east flow, governs the amount and timing of precipitation (rain and snow) and snowmelt in the high Sierra Nevada. Superimposed on the natural controls of freshwater flow in such a highly managed watershed as the Sacramento-San Joaquin system are the effects of human regulation of water flow represented by the storage, release, and export of freshwater at numerous points along the flow route between the mountains and the sea.
Management of the delta’s freshwater supply is complicated in part because California’s freshwater supply serves two masters: human consumption and salinity control. When freshwater runoff from the Sierra Nevada is scarce, as in dry years, a proportionally greater amount of available freshwater from the delta is needed for human use as well as for salinity control.
The extent to which seawater penetrates inland, particularly during summer, is a long-standing water management concern. About 20 million people depend on water flowing in the Sacramento-San Joaquin Delta inland from North Bay for their water supply. In addition, some aquatic species of the upper estuary are affected adversely by high salinity.
Perhaps the most critical period of the “water year” for meeting human needs, mostly agriculture, is in spring and summer during the peak period of farmland irrigation. Because of the adverse effects on fisheries associated with prolonged high salinity, legislation has been adopted that would limit salinity to 2 psu or less at Collinsville during the period February through June (a period that is critical to the success of several species). Historical episodes of salinity higher than 2 psu at Collinsville are mapped in fig. 2.
![Historical episodes of salinity](images/sal_fig2.gif)
Figure 2. Historical episodes of salinity higher than 2 practical salinity units at Collinsville are shown in the highlighted area as both observed and simulated values. The horizontal direction shows variation over the months of the year and the vertical direction shows variation from year to year. The match between simulated and observed values shows how closely this model mimics salinity variations found in the estuary: the pink area covers times when only the observed salinity was greater than 2; the light blue area covers times when only the simulated salinity was greater than 2; and the purple color covers the times when both observed and model simulations match for salinity above 2.
Compounding the water availability issue is the trend over the last three decades of increasing water resource demand, especially in spring (fig 3). Despite large saw-toothed fluctuations in the data reflecting interannual fluctuations in climate, there has been a long-term overall rise in spring export of freshwater inflow from the delta in response to growing demand for urban drinking water supplies and for irrigated agriculture. This rise is not caused solely by people. A slow variation in winter and spring climate has contributed a small part to the rise (see Dettinger and Cayan, 1995).
![Managed spring (March, April, and May) export of freshwater from the delta to supply a growing demand for farmland irrigation and urban drinking water](images/sal_fig3.gif)
Figure 3. Managed spring (March, April, and May) export of freshwater from the delta to supply a growing demand for farmland irrigation and urban drinking water.
Local and Delta Inflow Affect South Bay Salinity
In contrast to the situation in North Bay, much less freshwater enters South San Francisco Bay. South Bay receives freshwater from local streams, principally from Coyote, Guadalupe, and Alameda Creeks (fig. 1) only during winter, and from waste treatment facilities. The discharge from the creeks is small and is, thus, a minor contributor to the overall water budget of South Bay. As a result, circulation in the South Bay is generally considered more sluggish than in North Bay. However, its salinity is also influenced by water from both the coastal ocean and from North Bay especially during periods of high delta discharge. The north and south parts of the Bay, then, are connected through the central Bay where water from the coastal ocean, North Bay, and South Bay meet and mix.
South Bay is a scientific and management challenge because it features two sources of freshwater discharge (distant delta discharge and local streams) that play major roles in determining South Bay salinities. Differentiating effects of delta discharge from effects of local South Bay streams on salinity observations is difficult because the two sources often vary simultaneously. That is, when it rains in the Bay area, it generally rains regionally; thus, both the freshwater input from local sources and from the delta (North Bay) increase.
A key South Bay resource issue is the need to determine the sources of contaminants (distant or local). A distant source of contaminants is the discharge from the Sacramento and San Joaquin River Basins. A local source is the municipal waste discharged directly into South Bay. A first step in distinguishing local from distant contaminants is to distinguish local from distant freshwater. Another issue is the need to determine the mechanisms of contaminant dilution. For instance, if the contaminant concentration in South Bay increases when salinity decreases, then the freshwater source of decreasing salinity may also be the source of contamination. Both issues then require a knowledge of the temporal and spatial variations in freshwater sources in South Bay.
Simulation of Salinity in North Bay
Numerical models represent a quantitative way of accounting for and understanding the interdependent processes, from climatic variations to water resources management, that affect estuarine salintiy. A model also provides a convenient vehicle for testing options in managing resources. In this context, models may represent two different linkages: models that relate atmospheric circulation to stream discharge, and models that relate discharge to changes in patterns of estuarine salinity. Here we describe the latter, in order to link delta discharge to San Francisco Bay salinity over long time scales from days to decades.
![Spatial resolution of model](images/sal_fig4.gif)
Figure 4. For illustration, the results of a numerical simulation model that divides the entire San Francisco estuary into 50 segments (see inset) show that movement and mixing of water between segments is largely driven by delta inflow (in the inner estuary) and tidal forcing (using predicted values of astronomical variations in sea level height near Golden Gate, at the outer estuary). Thus, by computing the salinity of the segment for each day, the model accounts for fresh-water and saltwater inflow to each segment. The degree of success of the model can be gauged by comparing the model-computed values against an extensive history of salinity observations.
Because of the intense interest in the effects of salinity near the delta on agricultural and fisheries resources, a 27-year time series of daily-mean salinity at Collinsville was selected to compare with simulation results. This period, water years 1967 through 1993, encompasses a wide range of flood and drought conditions (fig. 5). The fit between observed and simulated salinity (fig. 5-b) provides encouragement that this model is useful in quantifying the salinity response to discharge variations over seasonal, interannual and decadal time scales. Such models can help to better define the climate/discharge/salinity connection over a range of flood and drought scenarios.
![Figure 5.A, estimated delta inflow into San Francisco Bay. B, observed and simulated mean-daily salinity at Collinsville](images/sal_fig5.gif)
Figure 5.A, estimated delta inflow into San Francisco Bay. B, observed and simulated mean-daily salinity at Collinsville. This model is useful in quantifying the salinity response to discharge over seasonal, interannual, and decadal time scales.
Simulation of Salinity in South Bay
By using a model to simulate the South Bay system numerically, we can illustrate the power of combining field observations with numerical models. In this example we attempt to separate delta-inflow effects from local South Bay inflow effects. As shown below, the delta influence is still the most important.
The numerical model can simulate salinity from a range of historical and hypothetical inflows. The response of the model, although not an exact representation of the real world, provides insight into expected and observed salinity responses in the estuary.
One approach in identifying the sources of freshwater is to illustrate four discharge scenarios of the South Bay salinity field: 1) salinity in response to negligible local and delta inflow (no inflow); 2) salinity in response to local inflow with negligible delta inflow (local inflow only); 3) salinity in response to delta inflow with negligible local inflow (delta inflow only); 4) salinity in response to both delta and local inflow.
Although the model simulates salinity for the entire estuary, we present only the results from the mouth of the estuary (near Bay Bridge), from the approximate mid-section (near San Mateo Bridge), and from the head of the South Bay (near Dumbarton Bridge) to compare with observations (fig. 6). For simplicity, these results include stream and waste discharge only at the south end of the South Bay, a major source of freshwater. The other smaller local streams along the bay have been neglected.
![Observed and simulated mean-daily salinity in South Bay](images/sal_fig6.gif)
Figure 6. Observed and simulated mean-daily salinity in South Bay, 1993 with differing simulated inflows. The vertical axis represents the salt content of the water. The horizontal axis shows how the salinity varies at each location (fig. 1) over the annual cycle in 1993. The blue areas are the differences between the model results and observed data for each of the four scenarios. Think of the blue part as model error and note how the blue is less from scenarios 1 to 4. The response of the model, although not an exact representation of the real world, provides insight into expected and observed responses in the estuary.
No Inflow: As expected, the no-inflow simulation (far left column fig. 6) bears little resemblance to observed salinities except during the first 3 months when discharge was low. Throughout the remainder of the year, simulated salinities are higher than observed.
Local Inflow Only: Again, agreement is fairly good for the first 3 months. Local streams play a major role in controlling salinity near the head of the estuary (see Dumbarton Bridge, second column from left side of fig. 6). Virtually none of the local-inflow “signal” reaches the Bay Bridge.
Delta Inflow Only: The example with high delta inflow and no local inflow (third column from left side of fig. 6) also shows, as expected, that much of the salinity variability near the mouth of South Bay (near the Bay Bridge) is controlled by delta inflow. Does some of the delta signal reach the Dumbarton Bridge?
Both Local and Delta Inflows: With both inflows, the simulated salinities track the observations well. Notably the salinity responses in mid-South Bay at the San Mateo Bridge suggest delta inflow is more of a controlling factor than local inflow. Even at the head of South Bay, at Dumbarton Bridge, the agreement between simulated and observed salinities improved by considering the effects of delta inflow in addition to local inflow.
Future Considerations
Syntheses of decades of climate and water flow data and the development and application of appropriate models are needed to fully understand and predict the complex relations among long-term climate patterns, regional and local precipitation and runoff patterns, and water use changes.
The combined application of long-term field observations and numerical modeling of estuarine salinity provides a foundation for monitoring and managing other more complex and less understood estuarine chemical properties such as waste-derived contaminants and plant nutrients. In essence, knowing how and why the salinity is changing is fundamental to knowing how and why the estuary and its resources are changing.
Acknowledgements
Delta discharge and export data were provided by Sheila Greene, California Department of Water Resources; salinity data at Collinsville were provided by Cheryl Bauman and Muriel Ferris, U.S. Bureau of Reclamation.
References
Caffrey, J.M., Cole, B.E., Cloern, J.E., Tyler, C. and Jassby, A., 1994, Studies of the San Francisco Bay, California, Estuarine Ecosystem Pilot Regional Monitoring Results, 1993: U.S. Geological Survey Open File Report 94-82, Menlo Park, CA, 411 p.
Cayan, D.R. and Peterson, D.H., 1993, Spring climate and salinity in the San Francisco Bay estuary: Water Resources Research 29, 293-303.
Cheng, R.T., Casulli, V. and Gartner, J.W., 1993, Tidal, residual, intertidal mud flat (TRIM) model and its applications San Francisco Bay, California: Estuarine, Coastal and Shelf Science 36, 235-280.
Dettinger, M.D. and Cayan D.R., 1995, Large-scale atmospheric forcing of recent trends toward early snowmelt runoff in California: Journal of Climate 8, 606-623.
Peterson, D.H. and Carlson, P., 1968, Influence of runoff on seasonal changes in salinity in San Francisco Bay, California: Fall meeting abstracts, Transactions American Geophysical Union 49, 704.
Peterson, D.H., Cayan, D.R., DiLeo, J., Noble, M. and Dettinger, M.D., 1995, The role of climate in estuarine variability: American Scientist 83, 58-67.
Uncles, R.J. and Peterson, D.H., 1995, A Microcomputer Model of Long-term Salinity in San Francisco Bay, Environment International, in press U.S. Geological Survey, 1994, Water Resources Data California Water Year 1993 Volume 2. Pacific Slope Basins from Arroyo Grande to Oregon State Line Except Central Valley: U.S. Geological Survey Water-Data Report CA-93-2, Sacramento, CA, 391 p.
Walters, R.A., Cheng, R.T., Conomos, T.J., 1985, Time scales of circulation and mixing processes of San Francisco Bay waters: In: Temporal Dynamics of an estuary: San Francisco Bay (eds, J. E. Cloern and F.H. Nichols). Developments in Hydrobiology, Dr. W. Junk publishers, 13-36.
Contributors: Daniel R. Cayan, Michael D. Dettinger, Jeanne Sandra DiLeo, Stephen E. Hager, Noah Knowles*, Frederic H. Nichols, Laurence E. Schemel, Richard E. Smith, Reginald E. Uncles **
For Further Information Contact:
David H. Peterson
U.S. Geological Survey
345 Middlefield Road MS 496
Menlo Park, CA 94025
(415) 354-3366 or dhpete@usgs.gov
**Institution for Marine Environmental Research
Prospect Place, The Hoe, Plymouth
Devon PL1 3DH ENGLAND
*Scripps Institution of Oceanography
University of California - San Diego
La Jolla, CA 92093
|