Salinity Intrusion in Northern San Francisco Bay: Observations and models

Lorenz G. Straub Award Ceremony and Seminar: Dr. Stephen Monismith, Obayashi Professor in the School of Engineering and Senior Fellow at the Woods Institute for the Environment, Stanford University

2013 Lorenz G. Straub Award Recipient: Esther Eke

The problem of predicting how the salinity field in estuaries responds to freshwater inflows is one that draws attention from both physical oceanographers and hydraulic engineers since it has both scientific and practical dimensions. In Northern San Francisco Bay, examination of 20+ years of data spanning the estuary shows that the overall structure of the salt field can be described using a single parameter, X2, the distance in km measured from the Golden Gate Bridge along the channel thalweg to where the salinity on the bottom is 2. Analysis of long-term monitoring of biological data (e.g. fish abundance) shows that much of ecological functioning of the estuary depends on X2 and so regulations have been developed specifying X2 position depending on time of year and hydrologic conditions. Because these regulations can require substantial amounts of water, it is necessary to efficiently predict the behavior of X2 with some accuracy so as to help manage the competing demands for California’s limited water supply.

In this talk, using several data sets including one that goes back to ca. 1960, I will discuss the observed behavior of X2 and how it responds to flow, Q. In general, the tendency of freshwater flows to carry salt out  of the estuary is balanced by the tendency of dispersion to move salt upstream. A surprising aspect of the X2-Q relation in Northern San Francisco Bay is that it is much weaker than would be inferred from classical estuarine circulation theory, behavior that we attribute to the effects of stratification on the turbulent flows that support upstream salt flux. I will present a rigorously derived but simplified model of salinity dynamics that can be used to understand this behavior and that can be used to create a dynamically based (rather than purely empirical) model of unsteady salinity intrusion. Finally, examination of the relevant data also suggests that inability to accurately measure freshwater flows during relatively dry periods may be a bigger limitation on accurate predictions of low-flow behavior than is choice of model structure.