UC Santa Barbara

The Southern California coastal region is one of the five areas in the world with Mediterranean climate. The long dry summer and cool wet winter not only attract a large population, but also create a unique hydrologic signature with strong temporal and spatial heterogeneity in soil moisture, streamflow and vegetation water use.. The sandy soil, fractured and uplifting bedrocks in undeveloped areas result in flashy and non-linear hydrologic responses to storms. Urbanization in this area further alters the hydrograph by changing the land cover and drainage patterns . These changes alter both water and nitrate fluxes, which may have substantial impacts on the downstream and coastal ecosystems, such as Great Kelp forest in the Santa Barbara channel. To understand and ultimately manage of water resources in this region must take these multiple factors – climate, soil and urbanization into account. Study and modeling of flow generation and nitrate transport mechanisms in the undeveloped area, and disturbance of nitrate cycling and nitrate export in the urban areas in this region may reveal new insight and improve the understanding of hydrologic processes, and provide sustainable watershed management strategies.

Southern California coastal mountain watersheds are characterized by a Mediterranean climate, sandy soil, shallow rock, flashy hydrologic responses and drought tolerant vegetation such as oak and chaparral. The geological condition in this area may fit the requirement condition of ‘fill & spill’ hypothesis of subsurface flow generation mechanism. We adapted ‘fill & spill’ submodel in an eco-hydrologic model, Regional Hydro-Ecologic Simulation system (RHESSys), implemented it in the semi-arid area, and compare its model performance with the traditional ‘continuous transmissivity’ model which assume the soil hydraulic conductivity follows a continuous exponential function. Our results show that in the recession period, the modeled discharge from ‘fill & spill’ model dropped much faster than the one from ‘continuous transmissivity’ model, and fit the observed discharge data better. ‘Fill & spill’ model is also less sensitive to small precipitation events, and tends to increased estimates of peak flow. Assessment of whether peak flow responses by the ‘Fill & spill’ model are a better representation of observed dynamics is challenging given that peak flow observed data is prone to uncertainties from different sources such as heterogeneity in precipitation distribution and errors in stream gauge stations, and may not reflect the true hydrologic response of watershed behavior. In summary, model assessments show that the ‘fill & spill’ model results in substantially different recession behavior and that, for the study watershed, the ‘fill & spill’ model may be a better predictor of the watershed hydrologic responses to precipitation events better than ‘continuous transmissivity’ model.

The long dry summer and cool wet winter of semi-arid climate in the undeveloped watersheds results in uneven vertical distribution of nitrate, with most of the nitrate concentrated at surface layer and limited in the deep soil. In our conceptual model, we showed that the vertical distribution of nitrate and its interaction with soil hydraulic conductivity controls the nitrate concentration-discharge patterns in the downstream. The uneven distribution of nitrate tends to produce an enrichment pattern, and the more evenly distributed nitrate will lead to a stable or weak dilution. In the eco-hydro model, we tested a combination of different hydraulic conductivity and vertical distribution of soil nitrate. Results show that at both patch and watershed scales, nitrate concentration-discharge patterns are sensitive to the hydraulic conductivity and vertical distribution of soil nitrate. In patch scale, modeling results match with the conceptual model. In watershed scale, because of the lateral replenishment of nitrate, concentration-discharge patterns are more complex and showing a 2-stage pattern with a transition point where patterns shift from enrichment to dilution. The hydraulic conductivity controls the transition point, and the interactions between conductivity and vertical nitrate distribution still affects the concentration-discharge pattern. This result is different from previous researches, which attributes the enrichment and dilution in the nitrate concentration-discharge relationship as the consequences of lateral expansion and contraction of saturated area. Our results indicate that, besides the analysis from horizontal hydrologic connectivity, the distribution of nitrate in the vertical direction and its interaction with the hydraulic conductivity also contributes to the nitrate concentration-discharge pattern.

The urbanization and expansion of impervious areas have great impact on the downstream nitrate concentration. Previous studies focus on the impact of impervious surface on collecting nitrate and other pollutant. How the water availability in vegetated area affecting the nitrate concentration in urban area is less well known and may be particularly important in semi-arid regions. Our results show that the total impervious area (TIA), effective impervious area (EIA), and vegetation types all have impacts on the downstream nitrate concentration. The TIA and EIA control the water availability in the vegetated land. The reduced TIA and EIA will increase the water availability in vegetated land and decrease the stream flow. As growth of vegetation and microbial activity in semi-arid areas is water limited, increased water availability can lead to more nitrate uptake by plant, but may also lead to increased nitrate released through soil decomposition processes. Results show that, the impact of the enhanced vegetation uptake is the dominant factor and consequently reducing TIA and EIA generally tends to reduce downstream nitrate flux and concentration. In some cases, however, the impact of a reducing EIA on reducing the stream discharge may offset the influence from the enhanced plant nitrate uptake and increase the nitrate concentration. These results suggest that improve water quality in urbanizing regions, a low EIA will generally but not always reduce nitrate concentration and the magnitude of this effect varies strongly with TIA and vegetation type. Thus the design of effective strategies may be improved by model-based assessments that account for interactions among EIA, TIA, and vegetation types. These findings are specific to semi-arid regions where additional water availability has great impacts. In humid environment where the biochemical activities are generally not water limited, some of the interactions in this study may not occur.