The California Department of Water Resources and Bureau of Reclamation propose new water intake facilities on the Sacramento River in northern California that would convey some of the water for export to areas south of the Sacramento-San Joaquin River Delta (hereinafter referred to as the Delta) through tunnels rather than through the Delta. The collection of water intakes, tunnels, pumping facilities, associated structures, and proposed operations are collectively referred to as California WaterFix. The water intake facilities, hereinafter referred to as the North Delta Diversion (NDD), are proposed to be located on the Sacramento River downstream of the city of Sacramento and upstream of the first major river junction where Sutter Slough branches from the Sacramento River. The NDD can divert a maximum discharge of 9,000 cubic feet per second (ft³/s) from the Sacramento River, which reduces the amount of Sacramento River inflow into the Delta.

In this report, we conducted three analyses to investigate the effect of the NDD and its proposed operation on entrainment of juvenile Chinook salmon (Oncorhynchus tshawytscha) into Georgiana Slough and the Delta Cross Channel (DCC). Fish that enter the interior Delta (the network of channels to the south of the Sacramento River) through Georgiana Slough and the DCC survive at lower rates than fish that use other migration routes (Sacramento River, Sutter Slough, and Steamboat Slough). Therefore, fisheries managers were concerned about the extent to which operation of the NDD would increase the proportion of the population entering the interior Delta, which, all else being equal, would lower overall survival through the Delta by increasing the fraction of the population subject to lower survival rates. Operation of the NDD would reduce flow in the Sacramento River, which has the potential to increase the magnitude and duration of reverse flows of the Sacramento River downstream of Georgiana Slough.

In the first analysis, we evaluate the effect of the NDD bypass rules on flow reversals of the Sacramento River downstream of Georgiana Slough. The NDD bypass rules are a set of operational criteria designed to minimize upstream transport of fish into Georgiana Slough and the DCC, and were developed based on previous studies showing that the magnitude and duration of flow reversals increase the proportion of fish entering Georgiana Slough and the DCC. We estimated the frequency and duration of reverse-flow conditions of the Sacramento River downstream of Georgiana Slough under each of the prescribed minimum bypass flows described in the NDD bypass rules. To accommodate adaptive levels of protection during different times of year when juvenile salmon are migrating through the Delta, the NDD bypass rules prescribe a series of minimum allowable bypass flows that vary depending on (1) month of the year, and (2) progressively decreasing levels of protection following a pulse flow event.

We determined that the NDD bypass rules increased the frequency and duration of reverse flows of the Sacramento River downstream of Georgiana Slough, with the magnitude of increase varying among scenarios. Constant low-level pumping, the most protective bypass rule that limits diversion to 10 percent of the maximum diversion and is implemented following a pulse-flow event, led to the smallest increase in frequency and duration of flow reversals. In contrast, we found that some scenarios led to sizeable increases in the fraction of the day with reverse flow. The conditions under which the proportion of the day with reverse flow can increase by greater than or equal to 10 percentage points between October and June, when juvenile salmon are present in the Delta, include October–November bypass rules and level-3 post-pulse operations during December–June. These conditions would be expected to increase the proportion of juvenile salmon entering the interior Delta through Georgiana Slough.

In the second analysis, we assessed bias in Delta Simulation Model 2 (DSM2) flow predictions at the junction of the Sacramento River, DCC, and Georgiana Slough. Because DSM2 was being used to simulate California WaterFix operations, understanding the extent of bias relative to USGS streamgages was important since fish routing models were based on flow data at streamgages. We determined that river flow predicted by DSM2 was biased for Georgiana Slough and the Sacramento River. Therefore, for subsequent analysis, we bias-corrected the DSM2 flow predictions using measured stream flows as predictor variables.

In the third analysis, we evaluated the effect of the NDD on the daily probability of fish entering Georgiana Slough and the DCC. We applied an existing model to predict entrainment from 15-minute flow simulations for an 82-year time series of flows simulated by DSM2 under the Proposed Action (PA), where the North Delta Diversion is implemented under California WaterFix, and the No Action Alternative (NAA), where the diversion is not implemented. To estimate the daily fraction of fish entering each river channel, entrainment probabilities were averaged over each day. To evaluate the two scenarios, we then compared mean annual entrainment probabilities by month, water year classification, and three different assumed run timings. Overall, the probability of remaining in the Sacramento River was lower under the PA scenario, but the magnitude of the difference was small (³/s. At flows greater than 41,000 ft³/s, we hypothesize that entrainment into the interior Delta is relatively constant, which would have caused little difference between scenarios at higher flows.