2.2.3. Models of Pre-Development Central Valley Hydrology and Delta Hydrodynamics

Pre-development Central Valley hydrology and Delta outflow were characterized by the California Department of Water Resources (CDWR) [25] utilizing two models to simulate watershed hydrology. They used the Soil Water Assessment Tool (SWAT) [57] to model precipitation–runoff characteristics of the upper elevation Central Valley watersheds and the (California) Central Valley Simulation Model, or C2VSim [58], an integrated hydrologic model, to simulate groundwater and surface water hydrology on the predevelopment Central Valley floor. Land use was based on prior characterizations of natural vegetation [24,59]. Potential evapotranspiration from natural vegetation was estimated using reference evapotranspiration from Orang et al. [60] and vegetation coefficients from Howes et al. [41]. CDWR [25] estimated a long-term annual average pre-development Delta outflow of 23.9 billion cubic meters (BCM) assuming a repeat of a 93-year contemporary climate sequence spanning WYs 1922–2014. Gross et al. [22] utilized these modeled values to compare inter- and intra-annual variability of pre-development and contemporary Delta outflow.

Pre-development salinity conditions in the San Francisco Estuary were investigated and compared to contemporary salinity conditions by Andrews et al. [26] using a threedimensional hydrodynamic model [61]. Their pre-development model was based on a planform developed by [62] and bathymetry from multiple sources. Their simulation used observed inflow data from February 2006 to October 2008 to represent wet, dry, and critically dry water years. Andrews et al. [26] found the dramatic changes in estuary planform and bathymetry, as well as differences in mean sea level between the pre-development and contemporary conditions, to have limited influence on saltwater intrusion. The predevelopment estuary was found to have less saltwater intrusion for the same Delta outflow and a faster response of saltwater intrusion to changes in Delta outflow. Due to the changes in seasonal distribution of Delta outflow, saltwater intrusion was found to be less variable for their contemporary scenario than their pre-development scenario. Changes to the seasonal timing of freshwater flows was reported to have a larger influence on saltwater intrusion than the changes in estuarine planform and bathymetry. Gross et al. [22] utilized this work to compare inter- and intra-annual variability of pre-development and contemporary salinity intrusion in the Delta.

The aforementioned model studies of pre-development conditions used an analysis method termed the "level-of-development" approach [63]. In this approach, landscape, channel geometry, and anthropogenic flow modification through reservoirs or withdrawals are fixed to represent a specific era or scenario (e.g., pre-development conditions, contemporary conditions, planned future conditions) and hydrology is typically represented by a sequence of historically observed precipitation or runoff. Thus, these model studies seek to describe how a pre-development or modern landscape and estuary would respond given contemporary instrumentally derived climatic inputs.

As described later in this paper, our work adopts some level-of-development assumptions to characterize pre-development Central Valley hydrology and Delta hydrodynamics. For example, we assume a stationary pre-development landscape consistent with Fox et al. [24] and CDWR [25] and a stationary pre-development outflow–salinity relationship consistent with Andrews et al. [26]. However, our work deviates from a typical level-of-development analysis in one crucial aspect—the driving hydrology is not simply

represented by repeating the sequence of observed runoff over the instrumented period. Rather, it reflects the estimated runoff over a millennial time scale obtained from the tree-ring proxy record from the watershed.
