Development, implementation, and validation of a modeling system for the San Francisco Bay and Estuary

Abstract

A three-dimensional numerical modeling system for the San Francisco Bay is presented. The system is based on an unstructured grid numerical model known as Semi-implicit Cross-scale Hydroscience Integrated System Model (SCHISM). The lateral boundary condition is provided by a regional coastal ocean model. The surface forcing is provided by a regional atmospheric model. The SCHISM results from a decadal hindcast run are compared with available tide gauge data, as well as a collection of temperature and salinity profiles.An examination of the observed climatological annual mean salinities at the United States Geological Survey (USGS) stations shows the highest salinities to be in the open ocean and the lowest well north (upstream) of the Central Bay, a pattern that does not change substantially with season. The corresponding mean SCHISM salinities reproduced the observed variations with location quite well, though with a fresh bias. The lowest values within the Bay occur during spring and the highest values during autumn, mirroring the seasonal variations in river discharge. The corresponding observed mean temperatures within the Bay were 2 to 3° C cooler in the Central Bay than to either the north or south. This observed pattern of a cooler Central Bay was not particularly well reproduced in the SCHISM results, which also showed a cold bias. Examination of the seasonal means revealed that the cool Central Bay pattern is found only during summer in the SCHISM results. The persistent cold and fresh biases in the model control run were nearly eliminated in a sensitivity run with modifications to the surface heat flux and river discharge.The surface atmospheric forcing and the heat flux at the western boundary are found to be the two major terms in a SCHISM-based heat budget analysis of the mean seasonal temperature cycle for the Central Bay. In the Central Bay salt budget, freshwater discharged by rivers into upstream portions of the Bay to the north balanced by the influx of salt from the west are the primary drivers of the mean seasonal salinity cycle. Concerning the interannual variability in temperatures, the warm anomalies during the period 2014–16 were the strongest and most persistent departures from normal during the period analyzed and were realistically reproduced by SCHISM. The most prominent salinity anomalies in both the observations and SCHISM results were the salty anomalies that persisted for most of the four-year California drought of 2012–2015.