THE LONG-TERM SALINITY FIELD IN SAN-FRANCISCO BAY

Data are presented on long-term salinity behaviour in San Francisco Ba y, California. A two-level, width averaged model of the tidally averag ed salinity and circulation has been written in order to interpret the long-term (days to decades) salinity variability. The model has been used to simulate daily averaged salinity in the upper and lower levels of a 51 segment discretization of the Bay over the 22-yr period 1967- 1988. Monthly averaged surface salinity from observations and monthly- averaged simulated salinity are in reasonable agreement. Good agree me nt is obtained from comparison with daily averaged salinity measured i n the upper reaches of North Bay. The salinity variability is driven p rimarily by freshwater inflow with relatively minor oceanic influence. All stations exhibit a marked seasonal cycle in accordance with the M editerranean climate, as well as a rich spectrum of variability due to extreme inflow events and extended periods of drought. Monthly averag ed salinity intrusion positions have a pronounced seasonal variability and show an approximately linear response to the logarithm of monthly averaged Delta inflow. Although few observed data are available for s tudies of long-term salinity stratification, modelled stratification i s found to be strongly dependent on freshwater inflow; the nature of t hat dependence varies throughout the Bay. Near the Golden Gate, strati fication tends to increase up to very high inflows. In the central rea ches of North Bay, modelled Stratification maximizes as a function of inflow and further inflow reduces stratification. Near the head of Nor th Bay, lowest summer inflows are associated with the greatest modelle d stratification. Observations from the central reaches of North Bay s how marked spring-neap variations in stratification and gravitational circulation, both being stronger at neap tides. This spring-neap varia tion is simulated by the model. A feature of the modelled stratificati on is a hysteresis in which, for a given spring-neap tidal range and f airly steady inflows, the stratification is higher progressing from ne aps to springs than from springs to neaps. The simulated responses of the Bay to perturbations in coastal sea salinity and Delta inflow have been used to further delineate the time-scales of salinity variabilit y. Simulations have been performed about low inflow, steady-state cond itions for both salinity and Delta inflow perturbations. For salinity perturbations a small, sinusoidal salinity signal with a period of 1 y r has been applied at the coastal boundary as well as a pulse of salin ity with a duration of one day. For Delta inflow perturbations a small , sinusoidally varying inflow signal with a period of 1 yr has been su perimposed on an otherwise constant Delta inflow, as well as a pulse o f inflow with a duration of one day. Perturbations in coastal salinity dissipate as they move through the Bay. Seasonal perturbations requir e about 40-45 days to propagate from the coastal ocean to the Delta an d to the head of South Bay. The response times of the model to perturb ations in freshwater inflow are faster than this in North Bay and comp arable in South Bay. In North Bay, time-scales are consistent with adv ection due to lower level, up-estuary transport of coastal salinity pe rturbations; for inflow perturbations, faster response times arise fro m both upper level, down-estuary advection and much faster, down-estua ry migration of isohalines in response to inflow volume continuity. In South Bay, the dominant time-scales are governed by tidal dispersion. Copyright (C) 1996 Elsevier Science Ltd