Abstract

A watershed model using Hydrologic Simulation Program—FORTRAN (HSPF) was developed for the urbanized Chino Basin in southern California to simulate the transport of pathogen indicator bacteria, evaluate the flow-component and land-use contributions to bacteria contamination and water-quality degradation throughout the basin, and develop a better understanding of the potential effects of climate and land-use change on water quality. The calibration of the model for indicator bacteria was supported by historical data collected before this study and by samples collected by the U.S. Geological Survey from targeted land-use areas during storms in water-year 2004. The model was successfully calibrated for streamflow at 5 gage locations representing the Chino Creek and Mill Creek drainages. Although representing pathogens as dissolved constituents limits the model’s ability to simulate the transport of pathogen indicator bacteria, the bacteria concentrations measured over the period 1998-2004 were well represented by the simulated concentrations for most locations. Hourly concentrations were more difficult to predict because of high variability in measured bacteria concentrations. In general, model simulations indicated that the residential and commercial land uses were the dominant sources for most of the pathogen indicator bacteria during low streamflows. However, simulations indicated that land used for intensive livestock (dairies and feedlots) and mixed agriculture contributed the most bacteria during storms.

The calibrated model was used to evaluate how various land use, air temperature, and precipitation scenarios would affect flow and transport of bacteria. Results indicated that snow pack formation and melt were sensitive to changes in air temperature in the northern, mountainous part of the Chino Basin, causing the timing and magnitude of streamflow to shift in the natural drainages and impact the urbanized areas of the central Chino Basin. The relation between bacteria concentrations and air temperature was more complicated, and did not substantially affect the quality of water discharging from the Chino Basin into the Santa Ana River. Changes in precipitation had a greater basin-wide affect on bacteria concentrations than changes in air temperature, and varied according to location. Drainages representing natural conditions had a decrease in bacteria concentrations in correlation with an increase in precipitation, whereas drainages in the central and southern part of the Chino Basin had an increase in bacteria concentrations. Drier climate conditions tended to result in higher sensitivity of simulated bacteria concentrations to changes in precipitation. Simulated bacteria concentrations in wetter climates were usually less sensitive to changes in precipitation because bacteria transport becomes more dependent on the land-use specified bacteria loading rates and the storage limits. Bacteria contamination from impervious-area runoff is affected to a greater degree by drier climates, whereas contamination from pervious-area runoff is affected to a greater degree by wetter climates. Model results indicated that the relation between precipitation, runoff, and bacteria contamination is complicated because after the initial bacteria washoff and transport from the land surfaces during the beginning of a storm period, subsequent runoff has fewer bacteria available for washoff, which then dilutes the concentrations of bacteria in the downstream reach. It was illustrated that pathogen indicator bacteria transport depends most significantly on the relation of imperviousness to runoff, which controls the frequency, and often the magnitude, of transport, and on the contribution of higher bacteria loading rates used for pervious land areas, especially intensive feedlots, to the infrequent, but very high, peaks of bacteria contamination.

The indicator bacteria transport model for the Chino Basin was based on the assumption that non-point bacteria loading rates can be defined according to 12 different land use categories. Results from water-quality sampling, model calibration, and model application indicated that important differences exist for loading rates and parameters controlling bacteria washoff between natural land use, urban land use, and agricultural land use. In addition, the fraction of impervious area for a given land use is a critical factor in determining the effect of storm runoff on downstream water quality. Increasing the impervious area usually increases the frequency of impaired water quality caused by bacteria that are washed off during smaller storms. An increase in the fraction of pervious area having higher non-point bacteria loading rates and washoff limits does not necessarily cause the frequency of impaired water quality to increase, and may even cause the frequency of impaired water quality to decrease, but the maximum for bacteria concentrations during the largest storms will likely be much higher. Additional sampling during large storms would likely provide a better assessment of non-point-source loading rates and washoff limits for pervious areas with agricultural and recreational land use that are likely sources of bacteria contamination.

First posted July 20, 2011